Radar signal processor

Description

BACKGROUND OF THE INVENTION

This invention relates to a signal processor of a vehicle-mounted array radar apparatus having a plurality of antenna components.

So far, an array radar apparatus for predicting presence or movement state of a preceding target, such as a vehicle, which exists in a forward direction in order to prevent a collision and maintain the inter-vehicle distance with respect thereto has been known as one of vehicle-mounted array radar apparatus.

Some vehicle-mounted array radar apparatus may have a plurality of observation means, a plurality of means for respectively extracting distance components, a plurality of means for respectively estimating correlation matrices with exponential smoothing, and means for estimating presence and movement state of target.

FIG. 2 is a block diagram showing a schematic structure of a conventional vehicle-mounted array radar apparatus, FIG. 3 is a view showing a structure of a FM-CW radar which is an instance of the observation means, and FIG. 4 is a view showing a structure of a pulse radar which is an instance of the observation means.

A conventional vehicle-mounted array radar apparatus as shown in FIG. 2 has observation means 1 through K which are denoted with reference numerals OB, means for respectively extracting distance components 1 through K which are denoted with reference numerals DS, means or respectively estimating correlation matrices with exponential smoothing 1 through N which are denoted with reference numerals RM, and means for estimating presence and movement state of target which is denoted with a reference numeral OM.

The observation means p (array component p=1, 2, . . . K) acquires an observation signal which includes information of presence and movement state of a target which is obtained from a transmitted signal which is a radar wave and a received signal which is received by an antenna. The movement states of a target means positions (distance and azimuth), velocities, accelerations and the like. As shown in FIG. 2, the observation means p (array component p=1, 2, . . . K) outputs observation signals XTp(t1), XTp(t2), XTp(t3), . . . XTp(tM) which are sampled at times t1, t2, . . . tM. M numbers of the observation signals which are sampled at times t1, t2, . . . tM is referred to as one snapshot. The observation means 1 through K correspond to array components 1 through K. A linear array has been known as a physical arrangement of the array components.

A case of a FM-CW radar is now exemplarily mentioned. A FM-CW radar which is the observation means OB has an oscillator 1, an transmitting amplifier 2, a transmitting antenna 3, a receiving antenna 5, a receiving amplifier 6, a distributor 7, a mixer 9, a filter 10, and an A/D converter 11, as shown in FIG. 3. The oscillator 1 oscillates millimetric-wave signals which are modulated such that the frequency increases and decreases linearly with passage of time to form a triangular wave form, and the millimetric-wave signal is emitted as a transmitted signal S1 via the transmitting amplifier 2 and the transmitting antenna 3. At the same time, the reflected wave of the transmitted signal is received as a received signal S2 through the receiving antenna 5. After amplifying the signal S2 by the receiving amplifier 6, it is mixed with the transmitted signal S1 by the distributor 7 and the mixer 9. The mixed signal S3 is filtered, and converted into a digital signal by the A/D converter 11 so as to be sampled. This sampled signal S3 is the observation signal outputted by the observation means OB. Although the array radar has a plurality of the observation means OB as shown in FIG. 2, the transmitted signals S1 in the respective observation means OB are common.

Next, a case of a pulse radar which is another instance of the observation means OB is now mentioned. The observation means OB of the pulse radar is comprised of the oscillator 1, the transmitting amplifier 2, the transmitting antenna 3, the receiving antenna 5, the receiving amplifier 6, the filter 10 and the A/D converter 11, as shown in FIG.4. The oscillator 1 oscillates pulses which are obtained by dividing a signal having high frequency f0 every equal interval at a cycle of fr[Hz], and the transmitted signal S1 is emitted through the transmitting amplifier 2 and the transmitting antenna 3 (pulse transmission). At the same time, the reflected wave of the transmitted signal S1 is received as a received signal S2 through the receiving antenna 5. After amplifying the signal S2 by the receiving amplifier 6, it is filtered, and is outputted to the A/D converter 11 through a phase detector (not shown). The signal is converted into the digital signal S3 by the A/D converter 11 which is triggered by pulse transmission, and the digital signal S3 is sampled. This sampling signal S3 is the observation signal S3 which the observation means OB outputs. Similar to the FM-CW radar, the transmitted signals S1 used in the respective observation means OB are common.

The means for extracting distance component p (array component p=1, 2, . . . K) as shown in FIG. 2 extracts distance components XRp(r1) , . . . XRp(rN) corresponding to distances r1, . . . rN from the observation signals XTp(t1), XTp(t2), XTp(t3), . . . XTp (tM) of one snapshot which are outputted from the observation means p.

For instance, a case of a FM-CW radar is now mentioned. When a target having velocity V exists at a position of distance r from the receiving antenna 5 (Correctly speaking, “distance r” is a half of a distance from the transmitting antenna up to the receiving antenna via a target. But, the distance r from the receiving antenna 5 is adopted as “distance r”, provided that the transmitting antenna and the receiving antenna are positioned at the same position. This explanation is applied to all descriptions in the present specification when referring to the distance up to a target r.), the observation signals XTp(t1), XTp(t2), XTp(t3), . . . XTp(tM) of the FM-CW radar include a frequency component of fB[Hz] as shown in expression (1).
$\begin{matrix} [Expression 1] \\ f_{B} = \frac{4 \cdot Δ F}{V_{C} \cdot T_{m}} \cdot r \pm \frac{2 \cdot F_{0}}{V_{C}} \cdot V [Hz] (+ : at the time of modulation by increasing frequency, - : at the time of modulation by decreasing frequency) & (1) \end{matrix}$

where r denotes distance to a target, V denotes relative velocity of a target, Vc is light speed, Δ F is frequency deviation width of frequency modulation, Tm denotes cycle period of frequency modulation, and Fo denotes central transmitting frequency. If relative velocity is neglected, the following relation which is shown by expression (2) is given between distance r and frequency fB.
$\begin{matrix} [Expression 2] \\ f_{B} = \frac{4 \cdot Δ F}{V_{C} \cdot T_{m}} \cdot r [Hz] & (2) \end{matrix}$

If respective frequency components fb corresponding to distances r1 through rN are obtained from the observation signals XTp(t1), XTp(t2), XTp(t3), . . . XTp(tM) which are time signal series, the thus obtained are distance components XRp(r1), . . . XRp(rN). If an absolute value of the distance component XRp(rq) is big, it is understood that a target exists at the distance rq. On the contrary, an absolute value of the distance component XRp(rq) is small, it is understood that a target does not exist at the distance rq. Fourier transformation or filtering with a band-pass filter is used as a method for extracting frequency components.

Besides, a case of a pulse radar is mentioned as another instance. If a target exists at the distance r, an echo from the target is observed in the observation signal at a time
$\begin{matrix} [Expression 3] \\ t_{e} = \frac{2}{V_{C}} \cdot r & (3) \end{matrix}$

from a time when emitting a pulse where r denotes distance to the target, and Vc denotes light speed. If the observation signals XTp(t1), XTp(t2), XTp(t3) . . . XTp(tM) are sampled at echo starting times te respectively corresponding to distances r1 through rN, the sampled are distance components XRp(r1), . . . XRp(rN). Such sampling may be conducted after simple sampling or average filtering.

The means for estimating correlation matrix with exponential smoothing q (distance rq:q=1, 2, . . . N) estimates a correlation matrix Rxxq which represents a correlation characteristics between array components (coherence) from output signals XR1(rq), XR2(rq), . . . XRK(rq) of the means for extracting distance component 1 through K, the output signals being signals relating to the distance rq which is obtained by each array component. The correlation matrix Rxxq is given by Expression (4). On this occasion, the correlation matrix Rxxq is a complex matrix with K rows and K columns, and a component with i-th row and j-th column is represented by rxxqij.
$\begin{matrix} [Expression 4] \\ R_{xxq} = [\begin{matrix} r_{xxq 11} & \dots & r_{xx q 1 j} & \dots & r_{xxq 1 K} \\ ⋮ & ⋮ & ⋮ \\ r_{xxqi 1} & \dots & r_{xxqij} & \dots & r_{xxqiK} \\ ⋮ & ⋮ & ⋮ \\ r_{xxqK 1} & \dots & r_{xxqKj} & \dots & r_{xxqKK} \end{matrix}] \overset{Δ}{=} [\begin{matrix} E [X R 1 (rq) X R 1 {(rq)}^{*} & \dots & E [XR 1 (rq) {XRj (rq)}^{*}] & \dots & E [XR 1 (rq) {XRK (rq)}^{*}] \\ ⋮ & ⋮ & ⋮ \\ E [XRi (rq) XR 1 {(rq)}^{*}] & \dots & E [XRi (rq) {XRj (rq)}^{*}] & \dots & E [XRi (rq) {XRK (rq)}^{*}] \\ ⋮ & ⋮ & ⋮ \\ E [XRK (rq) XR 1 {(rq)}^{*}] & \dots & E [XRK (rq) {XRj (rq)}^{*}] & \dots & E [XRK (rq) {XRK (rq)}^{*}] \end{matrix}] * represents a complex conjugate . E [] represents an operation for obtaining an expected value, and is actually estimated by time average in the unit of one snapshot . & (4) \end{matrix}$

Since the correlation matrix is estimated in snapshots, an estimated value of the correlation matrix after a first snapshot is represented by Rxxq(1), and an estimated value after a second snapshot is represented by Rxxq(2), . . . and an estimated value after a m-th snapshot is represented by Rxxq(m). When referring to the distance component in a specific snapshot, the distance rq component of the observation means p in the m-th snapshot is represented by XRp (rq, m), and similar expression is used for the observation signal, also.

With such kind of expression method, a correlation matrix observation value in the m-th snapshot Rtmp_xxq(m) is calculated by an expression (5).
$\begin{matrix} [Expression 5] \\ R_{tmp_xxq} (m) = [\begin{matrix} r_{tmp_xxq 11} (m) & \dots & r_{tmp_xxq 1 j} (m) & \dots & r_{tmp_xxq1K} (m) \\ ⋮ & ⋮ & ⋮ \\ r_{tmp_xxqi1} (m) & \dots & r_{rmp_xxqij} (m) & \dots & r_{tmp_xxqiK} (m) \\ ⋮ & ⋮ & ⋮ \\ r_{tmp_xxqK 1} (m) & \dots & r_{tmp_xxqKj} (m) & \dots & r_{tmp_xxqKK} \end{matrix}] \overset{Δ}{=} [\begin{matrix} XR 1 (rq, m) XR 1 {(rq, m)}^{*} & \dots & XR 1 (rq, m) {XRj (rq, m)}^{*} & \dots & XR 1 (rq, m) {XRK (rq, m)}^{*} \\ ⋮ & ⋮ & ⋮ \\ XRi (rq, m) XR 1 {(rq, m)}^{*} & \dots & XRi (rq, m) {XRj (rq, m)}^{*} & \dots & XRi (rq, m) {XRK (rq, m)}^{*} \\ ⋮ & ⋮ & ⋮ \\ XRK (rq, m) XR 1 {(rq, m)}^{*} & \dots & XRK (rq, m) {XRj (rq, m)}^{*} & \dots & XRK (rq, m) {XRK (rq, m)}^{*} \end{matrix}] & (5) \end{matrix}$

As a method of estimating a correlation matrix, a section average type and an exponential smoothing type are well-known.

In the method of estimating correlation matrix with section average, an average value between correlation matrix observation values of snapshots which continues predetermined SSN [times] is used as an estimated value of the correlation matrix. A number of sampling SSN [times] for equalization relates to S/N improvement. If SSN becomes bigger, influence of noise in an observation signal is removed, so that S/N improves. One of methods of calculating the estimated value is shown in an expression (6)
$\begin{matrix} [Expression 6] \\ R_{xxq} (m) = \frac{1}{SSN} \sum_{j = 0}^{SSN - 1} R_{tmp_xxq} (m - j) & (6) \end{matrix}$

Although the above-mentioned refers to a case where the estimated value of the correlation matrix is renewed, synchronizing with the snapshot, the renewal cycle of the estimated value of the correlation matrix may be once SSN snapshot times. In such a case, the estimated value of the correlation matrix is renewed as shown by an expression (7)
$[Expression 7]$ $?$ $? indicates text missing or illegible when filed$

where Rxxq(m′) means the estimated value of the correlation matrix which is renewed in the m′-th snapshot.

The method of estimating correlation matrix with exponential smoothing is one for obtaining the estimated value of the correlation matrix by respectively weighting the estimated value of the correlation matrix of the last snapshot and the observation value of the correlation matrix which is obtained in the present snapshot and adding both. A weight on the estimated value of the correlation matrix of the last snapshot is referred to as a forgetting factor, and is represented by α. At this time, a weight on the observation value of the correlation matrix which is obtained in the present snapshot is 1−α. A method of calculating the estimated value in the method of estimating correlation matrix with exponential smoothing is shown in an expression (8).

Expression 8

R_xxq(m)=α·R_xxy(m−1)+(1−a)·R_tmp_—_xxq(m) (8)

The number of sampling SSN [times] for equalization in the section average method which has been mentioned before and the forgetting factor α have the following relation as shown by an expression (9) in view of dispersion of the estimated value.
$\begin{matrix} [Expression 9] \\ α = \frac{SSN - 1}{SSN + 1} & (9) \end{matrix}$

The expression (9) is introduced by such a condition that dispersion of the estimated value is equal in the section average method and the exponential smoothing method if each element of the observation value of the correlation matrix conform to a chi-square distribution of degree of freedom 2, but this is not detailed mentioned. Then, SSN is made bigger as α approximates 1, so that the effect of the S/N improvement is made bigger. Therefore, the forgetting factor α is a parameter for adjusting the S/N improvement.

If the expression (8) is accepted as an IIR filter, a transient response performance on a change of the observation value of the correlation matrix is made better when α approximates zero (0), so that the forgetting factor α is a parameter for adjusting the transient response performance.

In the array radar apparatus as shown in FIG. 2, the means for estimating correlation matrix with exponential smoothing q (distance component rq: q=1, 2, . . . N) executes the estimating method of the expression (8), and the correlation matrix estimated value Rxxq corresponding to each distance component rq is outputted to the means for estimating presence and movement state of target OM.

The means for estimating presence and movement state of target OM estimates presence of a target and a movement state of the target from the correlation matrix estimated values Rxx1 through RxxN corresponding to the distances r1 through rN. FIG. 5 shows an instance of its schematic structure.

The means for estimating presence and movement state of target OM of FIG. 5 is comprised of means for computing angular spectrum corresponding to the distances r1 through rN which are denoted with reference numerals AS and data association means which is denoted with a reference numeral DA.

The means for computing angular spectrum q (distance rq:q=1, 2, . . . N) computes an angular spectrum Pq(θ) from the estimated value of the correlation matrix. The angular spectrum represents a power angular distribution, and Pq(θ1) represents a power which arrives from the azimuth θ1. Known computing methods are a beamformer method, a Capon's method, a Linear Prediction method, MUSIC, ESPRIT and the like (see non-patent-related document 1). In MUSIC or ESPRIT, the power can not be directly obtained, so it maybe referred to as a pseudo spectrum. But, all angular spectra including pseudo spectra are referred to as only angular spectra in the present embodiment. An intense reflected wave is considered to arrive from an angle where the spectral intensity is intense in the angular spectra, so that a target is predicted to be in this azimuth. In case of the angular spectrum Pq(θ) as shown in FIG. 6, for instance, a target is predicted to be at θ1, θ2 and θ3 [°] which spectrum intensities are intense.

The data association means DA estimates presence and a movement state of a target by inputting the angular spectra P1(θ) through PN(θ) corresponding to the distances r1 through rN. If the angular spectrum of FIG. 6 is inputted as one of the distance rq, for instance, the data association means DA estimates presence of three targets (distance, azimuth)=(rq, θ1), (distance, azimuth)=(rq, θ2) and (distance, azimuth)=(rq, θ3). If the angular spectrum of FIG. 7 is inputted as one of the distance rq, for instance, the data association means DA estimates no presence of target at the distance rq. The above-mentioned is a very simplified estimation instance. The positioning estimation accuracy may be increased by estimation with combination of the angular spectra P1(θ) through PN(θ) in a plurality of snapshots. Besides, the movement state, such as the velocity or the acceleration of the target, may be estimated in such a manner that a tracking process in an axial direction of the snapshot is applied to positioning information which is obtained from the angular spectrum.

In case of the FM-CW radar, the velocity of the target may be detected by Doppler effect by comparing two kinds of the angular spectra 1 through N, the snapshot at the time of modulation by increasing frequency and the snapshot at the time of modulation by decreasing frequency.

Besides, the movement state of the target may be known in such a way that the angular spectra 1 through N are overlapped with each other so as to prepare a two-dimensional distance-angular spectrum, and the prepared is treated as an image and is recognized by patterns (see non-patent-related document 2).

[Non-patent-related document 1] “Adaptive antenna technique” has been published on Oct. 10, 2003 by Ohmsha written by Nobuo KIKUMA

[Non-patent-related document 2] “Radar signal processing technique” has been published on Sep. 20, 1991 by “The Institute of Electronics, Information and Communication Engineers” written by Matsuo SEKINE, the chapter 10 “Pattern recognition in radar signal processing”.

The above-mentioned is the structure of the array radar signal processor for estimating the presence and the movement state of the target. In order to accurately estimate the presence and the movement state of the target, it is necessary to properly estimate the correlation matrix which is the intermediate data.

It is difficult to properly estimate the correlation matrix in the vehicle-mounted radar due to distance attenuation of the reflected wave from the target, a change of a reflected wave from a target which arises from a change of the movement state of ones own vehicle, such as the change in turning and the change of speed, a change of a reflected wave from a target which arises from a change of the surroundings of ones own vehicle, such as the change in the speed of the peripheral vehicle, a change of a radio wave state, such as an influence of an interfering wave.

The object of the invention is to provide the radar signal processor for properly estimating the correlation matrix according to the distance attenuation of the reflected wave from the target, the change of the reflected wave due to the changes of the movement state and surroundings of ones own vehicle, and the radio wave state.

BRIEF DESCRIPTION OF THE DRAWINGS

FIG. 1 is a block diagram showing a schematic structure of a vehicle-mounted array radar apparatus which is a first embodiment of the invention;

FIG. 2 is a block diagram showing a schematic structure of a conventional vehicle-mounted array radar apparatus;

FIG. 3 is a view showing a structure of a FM-CW radar which is an instance of observation means;

FIG. 4 is a view showing a structure of a pulse radar which is an instance of the observation means;

FIG. 5 is a block diagram showing a schematic structure of an instance of means for estimating presence and movement state of target;

FIG. 6 is a view showing an instance of an angular spectrum when there is target(s);

FIG. 7 is a view showing an instance of an angular spectrum when there is no target;

FIG. 8 is a block diagram showing a schematic structure of the vehicle-mounted array radar apparatus which is a second embodiment of the invention;

FIG. 9 is a view showing an instance of a target distance-input S/N graph;

FIG. 10 is a view showing routines for computing a forgetting factor α-azimuth RMSE graph;

FIG. 11 is a view showing an instance of a forgetting factor α-azimuth RMSE graph;

FIG. 12 is a view for explaining a method of determining the forgetting factor α according to distance attenuation;

FIG. 13 is a block diagram showing a schematic structure of the vehicle-mounted array radar apparatus which is a third embodiment of the invention;

FIG. 14 is a block diagram showing a schematic structure of the vehicle-mounted array radar apparatus which is a fourth embodiment of the invention;

FIG. 15 is a block diagram showing a schematic structure of the vehicle-mounted array radar apparatus which is a fifth embodiment of the invention;

FIG. 16 is a block diagram showing a schematic structure of the vehicle-mounted array radar apparatus which is a sixth embodiment of the invention;

FIG. 17 is a block diagram showing a schematic structure of the vehicle-mounted array radar apparatus which is a seventh embodiment of the invention;

FIG. 18 is a block diagram showing a schematic structure of the vehicle-mounted array radar apparatus which is a ninth embodiment of the invention;

FIG. 19 is a block diagram showing a schematic structure of the vehicle-mounted array radar apparatus which is a tenth embodiment of the invention;

FIG. 20 is a block diagram showing a schematic structure of the vehicle-mounted array radar apparatus which is an eleventh embodiment of the invention;

FIG. 21 is a block diagram showing a schematic structure of the vehicle-mounted array radar apparatus which is a thirteenth embodiment of the invention;

FIG. 22 is a block diagram showing a schematic structure of the vehicle-mounted array radar apparatus which is a fourteenth embodiment of the invention;

FIG. 23 is a block diagram showing a schematic structure of the vehicle-mounted array radar apparatus which is a fifteenth embodiment of the invention;

FIG. 24 is a block diagram showing a schematic structure of the vehicle-mounted array radar apparatus which is a seventeenth embodiment of the invention;

FIG. 25 is a block diagram showing a schematic structure of the vehicle-mounted array radar apparatus which is an eighteenth embodiment of the invention;

FIG. 26 is a block diagram showing a schematic structure of the vehicle-mounted array radar apparatus which is a nineteenth embodiment of the invention;

FIG. 27 is a block diagram showing a schematic structure of the vehicle-mounted array radar apparatus which is a twentyth embodiment of the invention; and

FIG. 28 is a block diagram showing a schematic structure of the vehicle-mounted array radar apparatus which is a twenty-second embodiment of the invention.

SUMMARY OF THE INVENTION

One aspect of the invention is radar signal processor, comprising:

- two or more observation means, each means for receiving a reflected wave from a target as a received signal through a receiving antenna and for outputting a predetermined observation signal;
- means for extracting distance component, for extracting a distance component corresponding to a distance from said receiving antenna from said each observation signal;
- means for estimating correlation matrix with exponential smoothing, for collecting respective said extracted distance components and for computing and estimating an estimated value of a correlation matrix corresponding to said respective distance components with a forgetting factor corresponding to said each distance component as a parameter by a method for estimating correlation matrix with exponential smoothing;
- means for estimating presence and movement state of target, for estimating a presence of a target or a movement state of said target on the basis of said respective estimated values of said correlation matrices which were estimated; and
- means for determining forgetting factor depending on distance, for respectively computing and determining said forgetting factors corresponding to said distances from said receiving antennas, and for respectively outputting said computed and determined forgetting factors as said forgetting factors corresponding to said respective distance components to said means for estimating correlation matrix with exponential smoothing.

According to this aspect of the invention, the forgetting factor α which is a parameter for the S/N improvement and the transient response in the estimation of a correlation matrix can be set, depending on distances. If a target at a distance (rq) is known in advance to strongly move, therefore, transient responsibility in estimation of a correlation matrix can be improved by setting the forgetting factor (α q) in the distance smaller than the other forgetting factors in order to follow the movement of the target, so that the correlation matrix can be properly estimated.

Another aspect of the invention is radar signal processor, comprising:

- two or more observation means, each means for receiving a reflected wave from a target as a received signal through a receiving antenna and for outputting a predetermined observation signal;
- means for extracting distance component, for extracting a distance component corresponding to a distance from said receiving antenna from said each observation signal;
- means for estimating correlation matrix with exponential smoothing, for collecting respective said extracted distance components and for computing and estimating an estimated value of a correlation matrix corresponding to said respective distance components with a forgetting factor corresponding to said each distance component as a parameter by a method for estimating correlation matrix with exponential smoothing;
- means for estimating presence and movement state of target, for estimating a presence of a target or a movement state of said target on the basis of said respective estimated values of said correlation matrices which were estimated; and
- means for determining forgetting factor depending on distance attenuation, for respectively computing and determining said forgetting factors corresponding to said distances from said receiving antennas according to an attenuation of said reflected wave due to said distance, and for respectively outputting said computed and determined forgetting factors as said forgetting factors corresponding to said respective distance components to said means for estimating correlation matrix with exponential smoothing.

According to this aspect of the invention, the forgetting factor α which is a parameter for the S/N improvement in the estimation of a correlation matrix can be made big according to the attenuation of reflected waves due to distance. Then, the influence of lowering of the S/N rate in the input of the reflected wave from the distant target due to the distance attenuation is improved, so that the correlation matrix can be properly estimated.

Another aspect of the invention is radar signal processor, comprising:

- two or more observation means, each means for receiving a reflected wave from a target as a received signal through a receiving antenna and for outputting a predetermined observation signal;
- range discrete Fourier Transform means, for respectively computing and extracting components of respective frequencies corresponding to said distance from said receiving antenna by discrete Fourier transform means on the basis of said observation signal;
- means for estimating correlation matrix with exponential smoothing, for collecting respective said extracted components of respective frequencies corresponding to said distances from said receiving antennas, and for computing and estimating an estimated value of a correlation matrix for said distance corresponding to said each frequency with a forgetting factor corresponding to said each distance as a parameter by a method for estimating correlation matrix with exponential smoothing;
- means for estimating presence and movement state of target, for estimating a presence of a target or a movement state of said target on the basis of said respective estimated values of said correlation matrices which were estimated; and
- means for determining forgetting factor depending on distance, for respectively computing and determining said forgetting factors corresponding to said distances from said receiving antennas, and for respectively outputting said computed and determined forgetting factors as said forgetting factors corresponding to said respective distance components to said means for estimating correlation matrix with exponential smoothing.

Another aspect of the invention is radar signal processor, comprising:

- two or more observation means, each means for receiving a reflected wave from a target as a received signal through a receiving antenna and for outputting a predetermined observation signal;
- range discrete Fourier Transform means, for respectively computing and extracting components of respective frequencies corresponding to said distance from said receiving antenna by discrete Fourier transform means on the basis of said observation signal;
- means for estimating correlation matrix with exponential smoothing, for collecting respective said extracted components of respective frequencies corresponding to said distance from said receiving antenna, and for computing and estimating an estimated value of a correlation matrix for said distance corresponding to said each frequency with a forgetting factor corresponding to said each distance as a parameter by a method for estimating correlation matrix with exponential smoothing;
- means for estimating presence and movement state of target, for estimating a presence of a target or a movement state of said target on the basis of said respective estimated values of said correlation matrices which were estimated; and
- means for determining forgetting factor depending on distance attenuation, for respectively computing and determining said forgetting factors corresponding to said distances from said receiving antennas according to an attenuation of said reflected wave due to said distance, and for respectively outputting said computed and determined forgetting factors as said forgetting factors corresponding to said respective distance components to said means for estimating correlation matrix with exponential smoothing.

According to this aspect of the invention, the forgetting factor α which is a parameter for the S/N improvement in the estimation of a correlation matrix can be made big according to the attenuation of reflected waves due to distance in a FM-CW radar. Then, the influence by lowering of the S/N rate in the input of the reflected wave from the distant target due to the distance attenuation is improved, so that the correlation matrix can be properly estimated.

Another aspect of the invention is radar signal processor, comprising:

- two or more observation means, each means for receiving a reflected wave from a target as a received signal through a receiving antenna and for outputting a predetermined observation signal;
- means for extracting distance component, for extracting a distance component corresponding to a distance from said receiving antenna from said each observation signal;
- means for estimating correlation matrix with section average, for collecting respective said extracted distance components and for computing and estimating a first estimated value of a correlation matrix corresponding to said respective distance components by a method for estimating correlation matrix with section average;
- means for estimating correlation matrix with exponential smoothing, for computing and estimating a second estimated value of a correlation matrix corresponding to said respective distance components with a forgetting factor corresponding to said each distance component as a parameter by a method for estimating correlation matrix with exponential smoothing from said first estimated value of said correlation matrix which was estimated;
- means for estimating presence and movement state of target, for estimating a presence of a target or a movement state of said target on the basis of said second estimated value of said correlation matrix which was estimated; and
- means for determining forgetting factor depending on distance, for respectively computing and determining said forgetting factors corresponding to said distances from said receiving antennas, and for respectively outputting said computed and determined forgetting factors as said forgetting factors corresponding to said respective distance components to said means for estimating correlation matrix with exponential smoothing.

According to this aspect of the invention, in the array radar apparatus which has means for estimating correlation matrix for estimating with section average and exponential smoothing in order, the forgetting factor α which is a parameter for the S/N improvement and the transient response in the estimation of a correlation matrix can be set, depending on distances.

Another aspect of the invention is radar signal processor, comprising:

- two or more observation means, each means for receiving a reflected wave from a target as a received signal through a receiving antenna and for outputting a predetermined observation signal;
- means for extracting distance component, for extracting a distance component corresponding to a distance from said receiving antenna from said each observation signal;
- means for estimating correlation matrix with exponential smoothing, for collecting respective said extracted distance components and for computing and estimating an estimated value of a correlation matrix corresponding to said respective distance components with a forgetting factor as a parameter by a method for estimating correlation matrix with exponential smoothing;
- means for estimating presence and movement state of target, for estimating a presence of a target or a movement state of said target on the basis of said respective estimated values of said correlation matrices which were estimated;
- a ones own vehicle sensor, for sensing a movement state of ones own vehicle and for outputting the sensed as ones own vehicle movement state information; and
- means for determining forgetting factor, for respectively computing and determining said forgetting factors on the basis of said ones own vehicle movement state information and for outputting said computed and determined forgetting factors to said means for estimating correlation matrix with exponential smoothing.

Another aspect of the invention is the radar signal processor, wherein said ones own vehicle sensor senses a turning state of ones own vehicle, and outputs ones own vehicle turning state information which was thus obtained as said ones own vehicle movement state information.

Another aspect of the invention is the radar signal processor, wherein said ones own vehicle sensor has means for judging turning/non-turning, for judging whether ones own vehicle is turning or not and for outputting a judged result as said ones own vehicle movement state information, and said means for determining forgetting factor has means for determining forgetting factor on turning state, for computing and determining said forgetting factor which has a low value if ones own vehicle is turning, and computing and determining said forgetting factor which has a high value if ones own vehicle is not turning on the basis of said ones own vehicle movement state information.

According to this aspect of the invention, a high transient responsibility in the estimation of a correlation matrix which is required at the time of turning can be achieved by making the forgetting factor α small, and a high S/N improvement in the estimation of a correlation matrix which is required at the time of non-turning can be achieved by making the forgetting factor α high, so that the correlation matrix can be properly estimated.

Another aspect of the invention is radar signal processor, comprising:

- two or more observation means, each means for receiving a reflected wave from a target as a received signal through a receiving antenna and for outputting a predetermined observation signal;
- means for extracting distance component, for extracting a distance component corresponding to a distance from said receiving antenna from said each observation signal;
- means for estimating correlation matrix with exponential smoothing, for collecting respective said extracted distance components and for computing and estimating an estimated value of a correlation matrix corresponding to said respective distance components with a forgetting factor as a parameter by a method for estimating correlation matrix with exponential smoothing;
- means for estimating presence and movement state of target, for estimating a presence of a target or a movement state of said target on the basis of said respective estimated values of said correlation matrices which were estimated;
- a steering sensor, for detecting an actual angle of a steering of ones own vehicle and for outputting the detected as ones own vehicle turning state information; and
- means for determining forgetting factor, for respectively computing arid determining said forgetting factors on the basis of said ones own vehicle turning state information, and for respectively outputting said computed and determined forgetting factors to said means for estimating correlation matrix with exponential smoothing.

According to this aspect of the invention, the forgetting factor α which is a parameter for the S/N improvement and the transient response in the estimation of a correlation matrix can be set according to a turning state of ones own vehicle with the steering sensor which can be easily mounted, so that the correlation matrix can be properly estimated according to the steering state.

Another aspect of the invention is radar signal processor, comprising:

- two or more observation means, each means for receiving a reflected wave from a target as a received signal through a receiving antenna and for outputting a predetermined observation signal;
- means for extracting distance component, for extracting a distance component corresponding to a distance from said receiving antenna from said each observation signal;
- means for estimating correlation matrix with exponential smoothing, for collecting respective said extracted distance components and for computing and estimating an estimated value of a correlation matrix corresponding to said respective distance components with a forgetting factor as a parameter by a method for estimating correlation matrix with exponential smoothing;
- means for estimating presence and movement state of target, for estimating a presence of a target or a movement state of said target on the basis of said respective estimated values of said correlation matrices which were estimated;
- a Yaw Rate sensor, for measuring a turning angle of ones own vehicle and for outputting the measured as ones own vehicle turning state information; and
- means for determining forgetting factor, for respectively computing and determining said forgetting factors on the basis of said ones own vehicle turning state information, and for respectively outputting said computed and determined forgetting factors to said means for estimating correlation matrix with exponential smoothing.

According to this aspect of the invention, the forgetting factor α which is a parameter for the S/N improvement and the transient response in the estimation of a correlation matrix can be set according to a turning state of ones own vehicle with the Yaw Rate sensor which can be easily mounted, so that the correlation matrix can be properly estimated according to the turning state of ones own vehicle.

Another aspect of the invention is radar signal processor, comprising:

- two or more observation means, each means for receiving a reflected wave from a target as a received signal through a receiving antenna and for outputting a predetermined observation signal;
- means for extracting distance component, for extracting a distance component corresponding to a distance from said receiving antenna from said each observation signal;
- means for estimating correlation matrix with exponential smoothing, for collecting respective said extracted distance components and for computing and estimating an estimated value of a correlation matrix corresponding to said respective distance components with a forgetting factor as a parameter by a method for estimating correlation matrix with exponential smoothing;
- means for estimating presence and movement state of target, for estimating a presence of a target or a movement state of said target on the basis of said respective estimated values of said correlation matrices which were estimated;
- a vehicle speed sensor, for sensing a vehicle speed of ones own vehicle and for outputting the sensed as vehicle speed information; and
- means for determining forgetting factor, for respectively computing and determining said forgetting factors on the basis of said vehicle speed information, and for respectively outputting said computed and determined forgetting factors to said means for estimating correlation matrix with exponential smoothing.

According to this aspect of the invention, the forgetting factor α which is a parameter for the S/N improvement and the transient response in the estimation of a correlation matrix can be set according to the vehicle speed of ones own vehicle, so that the correlation matrix can be properly estimated according to the speed of ones own vehicle.

Another aspect of the invention is the radar signal processor, wherein said vehicle speed sensor has means for judging vehicle speed, for judging whether ones own vehicle is running at a high speed or a low speed and for outputting the judged as said vehicle speed information, and said means for determining forgetting factor has means for determining forgetting factor on vehicle speed, for computing and determining said forgetting factor which has a small value if said vehicle speed which said vehicle speed information shows is high, and computing and determining said forgetting factor which has a high value if said vehicle speed which said vehicle speed information shows is low.

According to this aspect of the invention, a high transient responsibility in the estimation of a correlation matrix which is required at the time of high speed can be achieved by making the forgetting factor α small, and a high S/N improvement in the estimation of a correlation matrix which is required at the time of low speed can be achieved by making the forgetting factor α high, so that the correlation matrix can be properly estimated.

Another aspect of the invention is radar signal processor, comprising:

- two or more observation means, each means for receiving a reflected wave from a target as a received signal through a receiving antenna and for outputting a predetermined observation signal;
- means for extracting distance component, for extracting a distance component corresponding to a distance from said receiving antenna from said each observation signal;
- means for estimating correlation matrix with section average, for collecting respective said extracted distance components and for computing and estimating a first estimated value of a correlation matrix corresponding to said respective distance components by a method for estimating correlation matrix with section average;
- means for estimating correlation matrix with exponential smoothing, for computing and estimating a second estimated value of a correlation matrix corresponding to said respective distance components with a forgetting factor as a parameter by a method for estimating correlation matrix with exponential smoothing from said first estimated value of said correlation matrix which was estimated;
- means for estimating presence and movement state of target, for estimating a presence of a target or a movement state of said target on the basis of said second estimated value of said correlation matrix which was estimated;
- a ones own vehicle sensor, for sensing a movement state of ones own vehicle and for outputting the sensed as ones own vehicle movement state information; and
- means for determining forgetting factor, for respectively computing and determining said forgetting factors on the basis of said ones own vehicle movement state information and for outputting said computed and determined forgetting factors to said means for estimating correlation matrix with exponential smoothing.

Another aspect of the invention is radar signal processor, comprising:

- two or more observation means, each means for receiving a reflected wave from a target as a received signal through a receiving antenna and for outputting a predetermined observation signal;
- means for extracting distance component, for extracting a distance component corresponding to a distance from said receiving antenna from said each observation signal;
- means for estimating correlation matrix with exponential smoothing, for collecting respective said extracted distance components and for computing and estimating an estimated value of a correlation matrix corresponding to said respective distance components with a forgetting factor as a parameter by a method for estimating correlation matrix with exponential smoothing;
- means for estimating presence and movement state of target, for estimating a presence of a target or a movement state of said target on the basis of said respective estimated values of said correlation matrices which were estimated;
- means for obtaining surroundings of ones own vehicle, for obtaining surroundings of ones own vehicle and for outputting the obtained as information on surroundings of ones own vehicle; and
- means for determining forgetting factor, for respectively computing and determining said forgetting factors on the basis of said information on surroundings of ones own vehicle, and for respectively outputting said computed and determined forgetting factors to said means for estimating correlation matrix with exponential smoothing.

Another aspect of the invention is the radar signal processor, wherein said means for obtaining surroundings of ones own vehicle has means for obtaining information on speed limit, for obtaining a speed limit of a road where ones own vehicle is running and for outputting said obtained speed limit as said information on surroundings of ones own vehicle.

Another aspect of the invention is the radar signal processor, wherein said means for obtaining surroundings of ones own vehicle has means for judging speed limit, for judging as high speed limit if said speed limit of a road where ones own vehicle is running is higher than a predetermined value, and judging as low speed limit if said speed limit of said road where said own vehicle is running is lower than said predetermined value, and for outputting said judged high/low speed limit as said information on surroundings of ones own vehicle, and said means for determining forgetting factor has means for determining forgetting factor on speed limit, for computing and determining said forgetting factor which has a low value if said speed limit of said road where ones own vehicle is running is high, and computing and determining said forgetting factor which has a high value if said speed limit of said road where said ones own vehicle is running is low on the basis of said information on surroundings of ones own vehicle.

According to this aspect of the invention, a high transient responsibility in the estimation of a correlation matrix which is required for surroundings where the speed limit is high, such as a highway, can be achieved by making the forgetting factor α small, and a high S/N improvement in the estimation of a correlation matrix which is required for surroundings where the speed limit is low, such as an ordinary road, can be achieved by making the forgetting factor α high, so that the correlation matrix can be properly estimated.

Another aspect of the invention is the radar signal processor, wherein said means for obtaining information on speed limit has car navigation means for getting a present position of ones own vehicle, and for getting said speed limit of said present position from map information which is stored in advance.

According to this aspect of the invention, the forgetting factor α which is a parameter for the S/N improvement and the transient response in the estimation of a correlation matrix can be set according to the speed limit with the car navigation which can be easily mounted, so that the correlation matrix can be properly estimated.

Another aspect of the invention is radar signal processor, comprising:

- two or more observation means, each means for receiving a reflected wave from a target as a received signal through a receiving antenna and for outputting a predetermined observation signal;
- means for extracting distance component, for extracting a distance component corresponding to a distance from said receiving antenna from said each observation signal;
- means for estimating correlation matrix with section average, for collecting respective said extracted distance components and for computing and estimating a first estimated value of a correlation matrix corresponding to said respective distance components by a method for estimating correlation matrix with section average;
- means for estimating correlation matrix with exponential smoothing, for computing and estimating a second estimated value of a correlation matrix corresponding to said respective distance components with a forgetting factor as a parameter by a method for estimating correlation matrix with exponential smoothing from said first estimated value of said correlation matrix which was estimated;
- means for estimating presence and movement state of target, for estimating a presence of a target or a movement state of said target on the basis of said second estimated value of said correlation matrix which was estimated;
- means for obtaining surroundings of ones own vehicle, for obtaining surroundings of ones own vehicle and for outputting the obtained as information on surroundings of ones own vehicle; and
- means for determining forgetting factor, for respectively computing and determining said forgetting factors on the basis of said information on surroundings of ones own vehicle and for outputting said computed and determined forgetting factors to said means for estimating correlation matrix with exponential smoothing.

Another aspect of the invention is radar signal processor, comprising:

- two or more observation means, each means for receiving a reflected wave from a target as a received signal through a receiving antenna and for outputting a predetermined observation signal;
- means for extracting distance component, for extracting a distance component corresponding to a distance from said receiving antenna from said each observation signal;
- means for estimating correlation matrix with exponential smoothing, for collecting respective said extracted distance components and for computing and estimating an estimated value of a correlation matrix corresponding to said respective distance components with a forgetting factor as a parameter by a method for estimating correlation matrix with exponential smoothing;
- means for estimating presence and movement state of target, for estimating a presence of a target or a movement state of said target on the basis of said respective estimated values of said correlation matrices which were estimated;
- means for obtaining radio wave state, for obtaining radio wave state of surroundings of ones own vehicle from said received signal which was received by said receiving antenna, and for outputting the obtained as information on radio wave state; and
- means for determining forgetting factor, for respectively computing and determining said forgetting factors on the basis of said information on radio wave state, and for respectively outputting said computed and determined forgetting factors to said means for estimating correlation matrix with exponential smoothing.

Another aspect of the invention is the radar signal processor, wherein said means for obtaining radio wave state has means for computing interference, for computing and obtaining an interference influence due to waves excluding said signals transmitted from said observation means from said received signal, and for outputting said computed and obtained as said information on radio wave state.

Another aspect of the invention is the radar signal processor, wherein said means for determining forgetting factor has means for determining forgetting factor on interference influence, for computing and determining said forgetting factor as zero (0) if an interference influence in said information on radio wave state is higher than a predetermined threshold value.

According to this aspect of the invention, if the interference influence is specifically big, the estimated value of the correlation matrix which received the interference influence can be reset by making the forgetting factor α zero (0), so that the correlation matrix can be properly estimated.

Another aspect of the invention is radar signal processor, comprising:

- two or more observation means, each means for receiving a reflected wave from a target as a received signal through a receiving antenna and for outputting a predetermined observation signal;
- means for extracting distance component, for extracting a distance component corresponding to a distance from said receiving antenna from said each observation signal;
- means for estimating correlation matrix with section average, for collecting respective said extracted distance components and for computing and estimating a first estimated value of a correlation matrix corresponding to said respective distance components by a method for estimating correlation matrix with section average;
- means for estimating correlation matrix with exponential smoothing, for computing and estimating a second estimated value of a correlation matrix corresponding to said respective distance components with a forgetting factor as a parameter by a method for estimating correlation matrix with exponential smoothing from said first estimated value of said correlation matrix which was estimated;
- means for estimating presence and movement state of target, for estimating a presence of a target or a movement state of said target on the basis of said second estimated value of said correlation matrix which was estimated;
- means for obtaining radio wave state, for obtaining a radio wave state of surroundings of ones own vehicle from said received signal received by said receiving antenna and for outputting the obtained as information on radio wave state; and
- means for determining forgetting factor, for respectively computing and determining said forgetting factors on the basis of said information on radio wave state and for outputting said computed and determined forgetting factors to said means for estimating correlation matrix with exponential smoothing.

Another aspect of the invention is radar signal processor, comprising:

- two or more observation means, each means for receiving a reflected wave from a target as a received signal through a receiving antenna and for outputting a predetermined observation signal;
- means for estimating correlation matrix with exponential smoothing, for computing and estimating an estimated value of a correlation matrix corresponding to said observation signal on the basis of said each observation signal with a forgetting factor as a parameter by a method for estimating correlation matrix with exponential smoothing;
- means for estimating presence and movement state of target, for estimating a presence of a target or a movement state of said target on the basis of said respective estimated values of said correlation matrices which were estimated;
- a ones own vehicle sensor, for sensing a movement state of ones own vehicle and for outputting the sensed as ones own vehicle movement state information; and
- means for determining forgetting factor, for respectively computing and determining said forgetting factors on the basis of said ones own vehicle movement state information and for outputting said computed and determined forgetting factors to said means for estimating correlation matrix with exponential smoothing.

Another aspect of the invention is radar signal processor, comprising:

- two or more observation means, each means for receiving a reflected wave from a target as a received signal through a receiving antenna and for outputting a predetermined observation signal;
- means for estimating correlation matrix with section average, for computing and estimating a first estimated value of a correlation matrix corresponding to said observation signal on the basis of said each observation signal by a method for estimating correlation matrix with section average;
- means for estimating correlation matrix with exponential smoothing, for computing and estimating a second estimated value of a correlation matrix from said first estimated value of said correlation matrix which was estimated with a forgetting factor as a parameter by a method for estimating correlation matrix with exponential smoothing;
- means for estimating presence and movement state of target, for estimating a presence of a target or a movement state of said target on the basis of said second estimated value of said correlation matrix which was estimated;
- a ones own vehicle sensor, for sensing a movement state of ones own vehicle and for outputting the sensed as ones own vehicle movement state information; and
- means for determining forgetting factor, for respectively computing and determining said forgetting factors on the basis of said ones own vehicle movement state information and for outputting said computed and determined forgetting factors to said means for estimating correlation matrix with exponential smoothing.

Another aspect of the invention is radar signal processor, comprising:

- two or more observation means, each means for receiving a reflected wave from a target as a received signal through a receiving antenna and for outputting a predetermined observation signal;
- means for estimating correlation matrix with exponential smoothing, for computing and estimating an estimated value of a correlation matrix corresponding to said observation signal on the basis of said each observation signal with a forgetting factor as a parameter by a method for estimating correlation matrix with exponential smoothing;
- means for estimating presence and movement state of target, for estimating a presence of a target or a movement state of said target on the basis of said respective estimated values of said correlation matrices which were estimated;
- means for obtaining surroundings of ones own vehicle, for obtaining surroundings of ones own vehicle and for outputting the obtained as information on surroundings of ones own vehicle; and
- means for determining forgetting factor, for respectively computing and determining said forgetting factors on the basis of said information on surroundings of ones own vehicle and for outputting said computed and determined forgetting factors to said means for estimating correlation matrix with exponential smoothing.

Another aspect of the invention is radar signal processor, comprising:

- two or more observation means, each means for receiving a reflected wave from a target as a received signal through a receiving antenna and for outputting a predetermined observation signal;
- means for estimating correlation matrix with section average, for computing and estimating a first estimated value of a correlation matrix corresponding to said observation signal on the basis of said each observation signal by a method for estimating correlation matrix with section average;
- means for estimating correlation matrix with exponential smoothing, for computing and estimating a second estimated value of a correlation matrix from said first estimated value of said correlation matrix which was estimated with a forgetting factor as a parameter by a method for estimating correlation matrix with exponential smoothing;
- means for estimating presence and movement state of target, for estimating a presence of a target or a movement state of said target on the basis of said second estimated value of said correlation matrix which was estimated;
- means for obtaining surroundings of ones own vehicle, for obtaining surroundings of ones own vehicle and for outputting the obtained as information on surroundings of ones own vehicle; and
- means for determining forgetting factor, for respectively computing and determining said forgetting factors on the basis of said information on surroundings of ones own vehicle and for outputting said computed and determined forgetting factors to said means for estimating correlation matrix with exponential smoothing.

Another aspect of the invention is radar signal processor, comprising:

- two or more observation means, each means for receiving a reflected wave from a target as a received signal through a receiving antenna and for outputting a predetermined observation signal;
- means for estimating correlation matrix with exponential smoothing, for computing and estimating an estimated value of a correlation matrix corresponding to said observation signal on the basis of said each observation signal with a forgetting factor as a parameter by a method for estimating correlation matrix with exponential smoothing;
- means for estimating presence and movement state of target, for estimating a presence of a target or a movement state of said target on the basis of said respective estimated values of said correlation matrices which were estimated;
- means for obtaining radio wave state, for obtaining a radio wave state of surroundings of ones own vehicle from said received signal received from said receiving antenna and for outputting the obtained as information on radio wave state; and
- means for determining forgetting factor, for respectively computing and determining said forgetting factors on the basis of said information on radio wave state and for outputting said computed and determined forgetting factors to said means for estimating correlation matrix with exponential smoothing.

Another aspect of the invention is radar signal processor, comprising:

- two or more observation means, each means for receiving a reflected wave from a target as a received signal through a receiving antenna and for outputting a predetermined observation signal;
- means for estimating correlation matrix with section average, for computing and estimating a first estimated value of a correlation matrix corresponding to said observation signal on the basis of said each observation signal by a method for estimating correlation matrix with section average;
- means for estimating correlation matrix with exponential smoothing, for computing and estimating a second estimated value of a correlation matrix from said first estimated value of said correlation matrix which was estimated with a forgetting factor as a parameter by a method for estimating correlation matrix with exponential smoothing;
- means for estimating presence and movement state of target, for estimating a presence of a target or a movement state of said target on the basis of said second estimated value of said correlation matrix which was estimated;
- means for obtaining radio wave state, for obtaining a radio wave state of surroundings of ones own vehicle from said received signal received from said receiving antenna and for outputting the obtained as information on radio wave state; and
- means for determining forgetting factor, for respectively computing and determining said forgetting factors on the basis of said information on radio wave state and for outputting said computed and determined forgetting factors to said means for estimating correlation matrix with exponential smoothing.

Another aspect of the invention is the radar signal processor, wherein said means for obtaining surroundings of ones own vehicle has means for judging information on road division, for judging whether a road where ones own vehicle is running is a highway or an ordinary road, and for outputting a judged result as information on surroundings of ones own vehicle.

Another aspect of the invention is the radar signal processor, wherein said means for determining forgetting factor has means for determining forgetting factor on road kind, for computing and determining said forgetting factor which has a low value if said road where ones own vehicle is running is a highway, and computing and determining said forgetting factor which has a high value if said road where ones own vehicle is running is an ordinary road on the basis of said information on surroundings of ones own vehicle.

According to this aspect of the invention, a high transient responsibility in the estimation of a correlation matrix which is required for a highway where the speed limit is high can be achieved by making the forgetting factor α small, and a high S/N improvement in the estimation of a correlation matrix which is required for an ordinary road where the speed limit is low can be achieved by making the forgetting factor α high, so that the correlation matrix can be properly estimated.

Another aspect of the invention is radar signal processor, comprising:

- two or more observation means, each means for receiving a reflected wave from a target as a received signal through a receiving antenna and for outputting a predetermined observation signal;
- means for extracting distance component, for extracting a distance component corresponding to a distance from said receiving antenna from said each observation signal;
- means for estimating correlation matrix with exponential smoothing, for collecting respective said extracted distance components and for computing and estimating an estimated value of a correlation matrix corresponding to said respective distance components with a forgetting factor as a parameter by a method for estimating correlation matrix with exponential smoothing;
- means for estimating presence and movement state of target, for estimating a presence of a target or a movement state of said target on the basis of said respective estimated values of said correlation matrices which were estimated;
- means for setting target to be searched, for selectively setting a kind of said target to be searched, and for outputting a corresponding search kind signal; and
- means for determining forgetting factor, for respectively computing and determining said forgetting factors on the basis of said search kind signal, and for outputting said computed and determined forgetting factors to said means for estimating correlation matrix with exponential smoothing.

DESCRIPTION OF THE PREFERRED EMBODIMENTS

A radar signal processor 12 according to the invention has observation means 1 through K denoted with reference numerals OB, means for respectively extracting distance components 1 through K denoted with reference numerals DS, means for respectively estimating correlation matrices with exponential smoothing 1 through N denoted with reference numerals RM, means for estimating presence and movement state of target denoted with a reference numeral OM, and means for determining forgetting factor depending on distance denoted with a reference numeral FF, as shown in FIG. 1.

Explanations of the observation means p (array components p=1, 2, . . . K), the means for extracting distance component p (array components p=1, 2, . . . K) and means for estimating presence and movement state of target OM are omitted since these means are similar to ones of the conventional array radar apparatus as shown in FIG. 2.

The means for determining forgetting factor depending on distance FF independently determines forgetting factors α1, . . . α N, corresponding to distances r1, . . . rN. For instance, the means for determining forgetting factor depending on distance FF stores a table having indexes 1 through N in a memory means (not shown). A forgetting factor α N corresponding to distance rq is stored in index q in the table in advance, and the forgetting factor α q corresponding to distance rq is read out of the memory means so as to output to the means for estimating correlation matrix with exponential smoothing q corresponding to distance rq.

Expression 10

R_xxq(m)=α_q·R_xxq(m−1)+(1−α_q)·R_tmp_—_xxq(m) (10)

where Rxxq(m) denotes a correlation matrix estimated value after the m-th snapshot, Rxxq(m−1) denotes a correlation matrix estimated value after the (m−1)th snapshot, Rtmp_xxq(m) denotes a correlation matrix observation value in the m-th snapshot, and α q denotes the forgetting factor α for distance rq.

According to the invention, the forgetting factor α can be independently set, depending on distances. If a target at the distance rq is known to strongly move in advance, therefore, transient responsibility in estimation of a correlation matrix can be improved by setting the forgetting factor α q corresponding to the distance rq smaller than the other forgetting factors in order to follow the movement of the target.

Another embodiment of the invention is now mentioned. The radar signal processor 12 of FIG. 8 has the observation means 1 through K denoted with reference numerals OB, means for respectively extracting distance components 1 through K denoted with reference numerals DS, means for estimating correlation matrix with exponential smoothing 1 through N denoted with reference numerals RM, means for estimating presence and movement state of target denoted with a reference numeral OM, and means for determining forgetting factor depending on distance attenuation denoted with a reference numeral FF1.

Explanations of the observation means p (array components p=1, 2, . . . K), the means for extracting distance component p (array components p=1, 2, . . . K), the means for estimating correlation matrix with exponential smoothing q (distance rq:q=1, . . . N), and the means for estimating presence and movement state of target are omitted since these means are similar to ones of the embodiment of FIG. 1.

The means for determining forgetting factor depending on distance attenuation FF1 independently determines forgetting factors α1, . . . , α N corresponding to distances r1, . . . rN so as to match distance attenuation of reflective waves. For instance, the means for determining forgetting factor depending on distance attenuation FF1 stores a table having indexes 1 through N in memory means (not shown). The forgetting factor α N corresponding to distance rq is stored in index q in the table in advance, and the forgetting factor α q corresponding to distance rq is properly read out of the table in the memory means so as to output to the means for estimating correlation matrix with exponential smoothing q.

One of methods of determining forgetting factors α1 through α N, matching distance attenuation of reflected waves is shown below.

Received power of a reflected wave in a radar is modeled as shown in Expression (11) by a radar equation.
$\begin{matrix} [Expression 11] \\ P_{S} = \frac{P_{t} G^{2} λ^{2} σ}{{(4 π)}^{3}} \cdot \frac{1}{r^{4}} & (11) \end{matrix}$

where Ps denotes received power, Pt denotes sending power, G denotes antenna gain, λ denotes wave length of transmitted wave, σ denotes effective reflected area, and r denotes relative distance with a target. As seen from Expression (11), received power depends on distance, and monotonically attenuates. This is distance attenuation of received power.

At this time, input S/N to an antenna is modeled as shown by Expression (12).
$\begin{matrix} [Expression 12] \\ input S / N \overset{Δ}{=} \frac{P_{S}}{P_{N}} = \frac{P_{t} G^{2} λ^{2} σ}{{P_{N} (4 π)}^{3}} \cdot \frac{1}{r^{4}} & (12) \end{matrix}$

where PN denotes noise power due to thermal noise. Since thermal noise which may be noise is constant irrespective of distance to a target, input S/N (input S/N ratio) to an antenna also depends on distance, and monotonically attenuates. This is referred to as distance attenuation of input S/N. A target distance-input S/N graph of FIG. 9 which represents the influence on input S/N of distance attenuation is obtained with Expression (12). Although the above-mentioned refers to a case where the target distance-input S/N graph is obtained on the basis of the model of Expression (11), the method of obtaining on the basis of highly accurate model, taking consideration of ground reflected waves or the method of obtaining by actual measurement may be used.

Whether or not the estimated value of a correlation matrix is proper is judged by an estimated azimuth error which is an error between an estimated azimuth of a target which is obtained by computing an angular spectrum with the estimated value of the correlation matrix and the actual azimuth of the target. The influence of noise is stochastic, so that an azimuth RMSE (Root Mean Square Error) which is a two squares average of the estimated azimuth error is used as a standard for judging pertinence of the estimated value of the correlation matrix in this case.

The azimuth RMSE is obtained by a process as shown in FIG. 10. The process of FIG. 10 is simply explained. A pseudo signal generating means GS generates pseudo signals XR1 through XRK types of which are the same as the output signals of the means for extracting distance component DS, the output signals wherein a spectrum strength Ps of the angular spectrum P(θ) in angle θ t is Ps, and the spectrum strength is PN in the other angles, with the received power Ps, noise power PN and target azimuth θ t as parameters. The means for estimating correlation matrix with exponential smoothing RM estimates and outputs the correlation matrix estimated value Rxx with the forgetting factor α which is the parameter by inputting the pseudo signals XR1 through XRK by the exponential smoothing method. The angular spectrum computing means AS obtains and outputs the angular spectrum P(θ) with the correlation matrix estimated value Rxx by the method of computing angular spectrum, such as MUSIC. Means for computing RMSE of azimuth HC probes a peak of the angular spectrum P (θ), and obtains an estimated azimuth θ te of the target, and calculates |θ t−θ t| which is an error of the azimuth with the target. Such calculation of the error of the azimuth is repeated twice or more so as to obtain the average of the azimuth errors and output as the RMSE of the azimuth. If the process of FIG. 10 is used, the RMSE of the azimuth can be obtained when the forgetting factor α and the input SN(=Ps/PN) are two variables. FIG. 11 is a typical view showing a graph of forgetting factor α-azimuth RMSE.

The forgetting factor α corresponding to the distance rq is determined, matching the distance attenuation of the reflected wave with the graph of target distance−input S/N and the graph of forgetting factor α-azimuth RMSE as follows.

- A desired value of the azimuth RMSE which is a barometer of proper estimated value of the correlation matrix (desired azimuth RMSE) is set.
- The input S/N=(S/N)q corresponding to the target distance rq is obtained with the graph of target distance−input S/N (see left FIG. 12).
- A curve which corresponds to the input S/N=(S/N)q is selected with the graph of forgetting factor α-azimuth RMSE, and the forgetting factor α corresponding to the desired azimuth RMSE is determined as α q (see right FIG. 12).

The above-mentioned method of determining the forgetting factor is taken on all distances r1, . . . , rN so as to determine the forgetting factors α1, . . . , α N corresponding to the distance attenuation.

The above-mentioned method is one for determining the forgetting factors α1 through α N, matching the distance attenuation of the reflected waves. If the correlation matrix estimated values Rxx1 through Rxxq which are estimated with these forgetting factors α1 through α N are used, the means for estimating presence and movement state of target OM can compensate a constant azimuth RMSE in spite of the distance, controlling the influence of the distance attenuation.

Another embodiment of the invention is now explained. As shown in FIG. 13, the radar signal processor 12 of this embodiment has the observation means 1 through K denoted with reference numerals OB, range DFT (discrete Fourier transform) means 1 through K denoted with reference numerals KD, means for estimating correlation matrix with exponential smoothing 1 through N denoted with reference numerals RM, means for estimating presence and movement state of target denoted with a reference numeral OM, and means for determining forgetting factor depending on distance denoted with a reference numeral FF.

Explanations of the means for estimating correlation matrix with exponential smoothing q (distance rq:q=1, . . . , N), and the means for estimating presence and movement state of target OM, and the means for determining forgetting factor depending on distance FF are omitted since these means are similar to ones of the apparatus of FIG. 1.

In this embodiment, the radar wherein the frequency component of the observation signal obtained thereby corresponds to the distance component, such as the FM-CW radar, is used as the observation means p (array component p=1, 2, . . . K). When explaining with the case of the FM-CW radar, for instance, the observation signals XTp(t1), XTp(t2), XTp(t3), . . . , XTp(tM) include the frequency component of fB [Hz] as shown in Expression (13) if the relative velocity of the target is neglected.
$\begin{matrix} [Expression 13] \\ f_{B} = \frac{4 \cdot Δ F}{V_{C} \cdot T_{m}} \cdot r [Hz] & (13) \end{matrix}$

where r denotes the distance to the target, V denotes the relative velocity of the target, Vc denotes the light speed, Δ F denotes the frequency deviation width of frequency modulation, Tm denotes the cycle period of frequency modulation, and F₀denotes central transmitting frequency.

The range DFT (discrete Fourier transform) means p (array component P=1, 2, . . . K) computes and outputs frequency components XRp(f1), . . . , XRp(fN) corresponding to the frequencies fi, . . . fN by Discrete Fourier Transform from the observation signals S3. For instance, the frequency fq component XRp(fq) is calculated by Expression (14).
$\begin{matrix} [Expression 14] \\ {XR}_{p} (f_{q}) = \sum_{n = 1}^{N} XT (t_{i}) ⅇ^{- j \frac{2 π nq}{N}} & (14) \end{matrix}$

But, the frequency fq is defined by Expression (15)
$\begin{matrix} [Expression 15] \\ f_{q} \overset{Δ}{=} \frac{q}{N} \cdot \frac{1}{T_{s}} & (15) \end{matrix}$

if a sampling cycle of the observation signal is Ts. Since the frequency fq corresponds to the distance of Expression (16),
$\begin{matrix} [Expression 16] \\ r_{q} = \frac{V_{C} \cdot T_{m}}{4 \cdot Δ F} \cdot f_{q} & (16) \end{matrix}$

the frequency fq component XRp(fq) becomes the distance component in the distance rq which is defined by Expression (16).

The distance component of the distance rq which is defined by Expression (16) by the range DFT means q is outputted as the frequency component XRp(fq), so that subsequent process can be executed similarly to the embodiment of FIG. 1 and effects similar to the embodiment of FIG. 1 can be exercised.

Besides, Fast Fourier Transform can be used in the computing with Expression (14) for speedy computing.

Another embodiment of the invention is now explained, referring to FIG. 14. The radar signal processor 12 of FIG. 14 has the observation means 1 through K denoted with reference numerals OB, range DFT means 1 through K denoted with reference numerals KD, means for estimating correlation matrix with exponential smoothing 1 through N denoted with reference numerals RM, means for estimating presence and movement state of target denoted with a reference numeral OM, and means for determining forgetting factor depending on distance attenuation denoted with a reference numeral FF1.

Explanations of the means for estimating correlation matrix with exponential smoothing q (distance rq:q=1, . . . , N), and the means for estimating presence and movement state of target, and the means for determining forgetting factor depending on distance attenuation are omitted since these means are similar to ones of the invention of FIG. 8.

Besides, explanations of the observation means p (array components P=1, 2, . . . , K), and the range DFT means p (array components P=1, 2, . . . , K) are omitted since these means are similar to ones of the embodiment of FIG. 13.

According to this invention, in the array radar apparatus having the radar wherein the frequency component of the observation signal S3 obtained by the observation means OB corresponds to the distance component, such as the FM-CW radar as the observation means, the means for estimating presence and movement state of target OM can compensate a constant azimuth RMSE in spite of the distance, controlling the influence of the distance attenuation.

Another embodiment of the invention is now mentioned, referring to FIG. 15. The radar signal processor 12 of FIG. 15 has the observation means 1 through K denoted with reference numerals OB, means for respectively extracting distance components 1 through K denoted with reference numerals DS, means for estimating correlation matrix with section average 1 through N denoted with reference numerals KS, means for estimating correlation matrix with exponential smoothing 1 through N denoted with reference numerals RM, means for estimating presence and movement state of target denoted with a reference numeral OM, and means for determining forgetting factor depending on distance denoted with a reference numeral FF.

Explanations of the observation means p (array components P=1, 2, . . . K), and the means for extracting distance component p (array components P=1, 2, . . . , K), the means for estimating presence and movement state of target OM, the means for determining forgetting factor depending on distance FF are omitted since these means are similar to ones of the embodiment of FIG. 1.

The means for estimating correlation matrix with section average q (distance rq, q=1, 2, . . . , N) calculates the correlation matrix observation value Rtmp_xxq(m) of Expression (5) from the distance rq components XR1(rq), . . . , XRK(rq), and computes estimation of the correlation matrix with section average by Expression (6) or (7), and the computed is outputted as the first correlation matrix estimated value Rxxq_1(m).

The means for estimating correlation matrix with exponential smoothing q (distance rq: q=1, 2, . . . , N) uses the first correlation matrix estimated value Rxxq_1(m) in place of the observation value of the correlation matrix, and estimates the second correlation matrix estimated value Rxxq_2(m) by the exponential smoothing method and outputs this. Such estimation is executed by Expression (17).

Expression 17

R_xxq_—₂(m)=α_q·R_xxq_—₂(m−1)+(1−α_q)·R_xxq_—₁(m) (17)

In the array radar apparatus for estimating the correlation matrix with the means for estimating correlation matrix with section average KS and the means for estimating correlation matrix with exponential smoothing RM which are arranged in series, the forgetting factors α can be independently set according to the distance.

Another embodiment of the invention is now mentioned, referring to FIG. 16. The radar signal processor 12 of FIG. 16 has the observation means 1 through K denoted with reference numerals OB, means for respectively extracting distance components 1 through K denoted with reference numerals DS, means for estimating correlation matrix with exponential smoothing 1 through N denoted with reference numerals RM, means for estimating presence and movement state of target denoted with a reference numeral OM, a ones own vehicle sensor denoted with a reference numeral SN, and means for determining forgetting factor denoted with a reference numeral FX.

Explanations of the observation means p (array components P=1, 2, . . . K), and the means for extracting distance component p (array components P=1, 2, . . . K), means for estimating correlation matrix with exponential smoothing q (distance rq: q=1, 2, . . . , N), the means for estimating presence and movement state of target OM are omitted since these means are similar to ones of the embodiment of FIG. 1.

The ones own vehicle sensor SN senses a movement state of ones own vehicle, and outputs information obtained thereby as ones own vehicle movement state information AI. The ones own vehicle movement state information AI are a velocity of ones own vehicle Vm, an acceleration of ones own vehicle am, a turn angle velocity γ, a turning radius ρ, and the like. For instance, the ones own vehicle sensor SN obtains the above-mentioned ones own vehicle movement state information AI by getting a track of ones own vehicle by a tracking process with positioning information which is obtained by a sensor for obtaining positioning information, such as a GPS.

The means for determining forgetting factor FX determines the forgetting factors α to be given to the means for estimating correlation matrix with exponential smoothing 1 through N on the basis of the ones own vehicle movement state information AI. For instance, the forgetting factor α is determined in such a way that a table having acceleration of ones own vehicle am as an index is stored in a memory (not shown), and the corresponding forgetting factor α is read out of the memory according to the acceleration of ones own vehicle am which is obtained as the ones own vehicle movement state information AI, referring to the table.

The optional forgetting factor α can be thus set according to the movement state of ones own vehicle.

Another embodiment of the invention is now mentioned, referring to FIG. 17. The radar signal processor 12 of FIG. 17 has the observation means 1 through K denoted with reference numerals OB, means for respectively extracting distance components 1 through K denoted with reference numerals DS, means for respectively estimating correlation matrices with exponential smoothing 1 through N denoted with reference numerals RM, means for estimating presence and movement state of target denoted with a reference numeral OM, the ones own vehicle sensor denoted with a reference numeral SN, and means for determining for getting factor denoted with the reference numeral FX.

Explanations of the observation means p (array components P=1, 2, . . . K), and the means for extracting distance component p (array components P=1, 2, . . . K), means for estimating correlation matrix with exponential smoothing q (distance rq: q=1, . . . N), the means for estimating presence and movement state of target OM are omitted since these means are similar to ones of the embodiment of FIG. 1.

The ones own vehicle sensor SN senses a turning state of ones own vehicle, and outputs information obtained thereby as ones own vehicle turning state information AI1. The ones own vehicle turning state information AI1 are the turn angle velocity γ, a turning radius ρ, and the like. The sensor similar to one of the embodiment of FIG. 16 is used as the ones own vehicle sensor SN.

The means for determining forgetting factor FX determines the forgetting factors α to be given to the means for estimating correlation matrix with exponential smoothing 1 through N on the basis of the ones own vehicle turning state information AI1. For instance, the forgetting factor α is determined in such a way that a table having the turn angle velocity γ as an index is stored in memory means (not shown), and the corresponding forgetting factor α is read out of the memory according to the turn angle velocity γ which is obtained as the ones own vehicle movement state information AI, referring to the table.

The optional forgetting factor ct can be thus set according to the turning state of ones own vehicle.

Besides, in FIG. 17, the means for determining forgetting factor FX may judge as to whether ones own vehicle turns or not on the basis of the ones own vehicle turning state information AI1, and determine the forgetting factors α to be given to the means for estimating correlation matrix with exponential smoothing 1 through N according to the result of the judgment. Whether ones own vehicle turns or not is judged by the turn angle velocity γ, for instance. Non-turning (including straight going) is judged if the turn angle velocity γ is smaller than a predetermined threshold value γ , and turning is judged if the turn angle velocity γ is a predetermined threshold value γ or more. On the basis of such a judgment, turning or non-turning, the forgetting factor α is selected and outputted in such a way that the small forgetting factor α of the most suitable predetermined ones is read out of the memory means at the time of turning, and the big forgetting factor α of the most suitable predetermined ones is read out of the memory means at the time of non-turning.

It is thus possible to actualize high transient response performance of the correlation matrix estimation which is required for a change of the target signal at the time of turning and high S/N improvement which is required at the time of non-turning.

Another embodiment of the invention is now mentioned, referring to FIG. 18. The radar signal processor 12 of FIG. 18 has a steering sensor SN1 in place of the ones own vehicle sensor SN in the embodiment of FIG. 17, so that explanations of the structural elements excluding the steering sensor SN1 are omitted.

The steering sensor SN1 observes an actual steering angle δ of ones own vehicle. There is the following relation as shown by Expression (18) between the actual steering angle δ and the turn angle velocity γ.
$\begin{matrix} [Expression 18] \\ γ = \frac{V_{m}}{l} \cdot δ & (18) \end{matrix}$

where Vm denotes a velocity of ones own vehicle, and 1 denotes a wheel base. Then, the turn angle velocity γ is proportional to the actual steering angle δ provided that Vm is constant. Then, the steering sensor outputs the actual steering angle δ as the ones own vehicle turning state information AI1 in place of the turn angle velocity γ.

Effects similar to ones in the embodiment of FIG. 17 can be thus exercised with the steering sensor which is easily installed.

Another embodiment of the invention is now mentioned, referring to FIG. 19. The radar signal processor 12 of FIG. 19 has a Yaw Rate sensor SN2 in place of the ones own vehicle sensor SN of FIG. 17, so that explanations of the structural elements excluding the Yaw Rate sensor SN2 are omitted.

The Yaw Rate sensor SN2 directly observes the turn angle velocity γ with a gyro mechanism, and outputs the turn angle velocity γ as the ones own vehicle turning state information AI1.

Effects similar to ones in the invention of FIG. 17 can be thus exercised with the Yaw Rate sensor which is easily installed.

Another embodiment of the invention is now mentioned, referring to FIG. 20. The radar signal processor 12 of FIG. 20 has the observation means 1 through K denoted with reference numerals OB, means for respectively extracting distance components 1 through K denoted with reference numerals DS, means for respectively estimating correlation matrices with exponential smoothing 1 through N denoted with reference numerals RM, means for estimating presence and movement state of target denoted with the reference numeral OM, a vehicle speed sensor denoted with a reference numeral SN3, and means for determining forgetting factor denoted with the reference numeral FX.

Explanations of the observation means p (array components P=1, 2, . . . K), and the means for extracting distance component p (array components P=1, 2, . . . K) means for estimating correlation matrix with exponential smoothing q (distance rq: q=1, . . . , N), the means for estimating presence and movement state of target OM are omitted since these means are similar to ones of the embodiment of FIG. 1.

The vehicle speed sensor SN3 senses a vehicle speed Vm of ones own vehicle, and outputs the sensed as vehicle speed information AI2.

The means for determining forgetting factor FX determines the forgetting factors α to be given to the means for estimating correlation matrix with exponential smoothing 1 through N on the basis of the vehicle speed information AI2. For instance, the forgetting factor α is determined in such a way that a table having the vehicle speed Vm as an index is stored in memory means, and the forgetting factor α is read out of the memory means according to the vehicle speed Vm which is obtained as the vehicle speed information AI2, referring to the table.

The optional forgetting factor α can be thus set according to the vehicle speed.

Besides, the means for determining forgetting factor FX may judge high speed or low speed in such a way that high speed is judged if the vehicle speed of ones own vehicle is higher than a predetermined threshold value V0, and low speed is judged if the vehicle speed is a predetermined threshold value V0 or lower. On the basis of such a judgment, high speed or low speed, the small forgetting factor α is selected from the most suitable predetermined ones at the time of high speed, and the big forgetting factor α is selected from the most suitable predetermined ones at the time of low speed, and the selected is outputted.

It is thus possible to properly actualize high transient response performance of the correlation matrix estimation which is required due to increase of relative speed with a target which is stationary or travels at low speed when ones own vehicle travels at high speed and high S/N improvement which is required at the time of low speed.

Another embodiment of the invention is now mentioned, referring to FIG. 21. The radar signal processor 12 of FIG. 21 has the observation means 1 through K denoted with reference numerals OB, means for respectively extracting distance components 1 through K denoted with reference numerals DS, means for estimating correlation matrix with section average 1 through N denoted with reference numerals KS, the means for respectively estimating correlation matrices with exponential smoothing 1 through N denoted with reference numerals RM, means for estimating presence and movement state of target denoted with the reference numeral OM, the ones own vehicle sensor denoted with the reference numeral SN, and the means for determining forgetting factor denoted with the reference numeral FX.

Explanations of the observation means p (array components P=1, 2, . . . K), and the means for extracting distance component p (array components P=1, 2, . . . K), the means for estimating correlation matrix with section average q (distance rq: q=1, . . . , N), the means for estimating correlation matrix with exponential smoothing q (distance rq: q=1, . . . , N), the means for estimating presence and movement state of target OM are omitted since these means are similar to ones of the embodiment of FIG. 15.

Besides, explanations of the ones own vehicle sensor SN and the means for determining forgetting factor FX are omitted since these means are similar to ones of the embodiment of FIG. 16.

In the array radar apparatus for estimating the correlation matrix with the means for estimating correlation matrix with section average KS and the means for estimating correlation matrix with exponential smoothing RM which are provided in series, the forgetting factors α can be thus optionally set according to the movement state of ones own vehicle.

Another embodiment of the invention is now mentioned, referring to FIG. 22. The radar signal processor 12 of FIG. 22 has the observation means 1 through K denoted with reference numerals OB, means for respectively extracting distance components 1 through K denoted with reference numerals DS, the means for respectively estimating correlation matrices with exponential smoothing 1 through N denoted with reference numerals RM, the means for estimating presence and movement state of target denoted with the reference numeral OM, means for obtaining surroundings of ones own vehicle denoted with a reference numeral SN4, and the means for determining forgetting factor denoted with the reference numeral FX.

Explanations of the observation means p (array components P=1, 2, . . . K), and the means for extracting distance component p (array components P=1, 2, . . . K), the means for estimating correlation matrix with exponential smoothing q (distance rq: q=1, . . . , N), the means for estimating presence and movement state of target OM are omitted since these means are similar to ones of the embodiment of FIG. 1.

The means for obtaining surroundings of ones own vehicle SN4 obtains the surroundings of ones own vehicle, and outputs the obtained as information on surroundings of ones own vehicle AI3. The surroundings are whether a target exist or not in the surroundings, a shape of a road, speed limit, a division of a road, highway or ordinary road, and the like. As a method of obtaining the surroundings of ones own vehicle, a peripheral image is obtained with a CCD sensor, and by picture processing and/or picture recognition on the obtained image, the presence of a target in the surroundings, the shape of a road, a speed limit, and a division of a road are known.

The means for determining forgetting factor FX determines the forgetting factors α to be given to the means for estimating correlation matrix with exponential smoothing 1 through N on the basis of the information on surroundings of ones own vehicle AI3. For instance, in case of no presence of a target, the smaller forgetting factor α is selected in order to improve a transient response in preparation for the presence of a new target, and in case of the presence of a target, the big forgetting factor α is selected in order to continuously watch the target according to the presence of a target in the surroundings. By doing so, the forgetting factor a can be optionally set according to the surroundings of ones own vehicle.

Besides, the forgetting factor α which is determined by the means for determining forgetting factor FX may be changed according to the target which is searched by the radar signal processor 12, such as a pedestrian, a vehicle or a bicycle which suddenly appears from a side road, and a preceding vehicle in the forwarding direction.

In such a case, the means for determining forgetting factor FX determines the small forgetting factor α in order to improve a transient response for the target, the relative moving speed of which is high on ones own vehicle, such as a pedestrian, a bicycle and a vehicle which suddenly rush into the forwarding direction, and determines the big forgetting factor α in order to lower a transient response for the target, the relative moving speed of which is low on ones own vehicle, such as a preceding vehicle in the forwarding direction wherein a change of inter-vehicle distance to ones own vehicle is small.

Concretely speaking, kinds of the targets to be searched may be properly set by operating means for setting target to be searched, for inputting kinds of the targets to be searched, such as a switch (not shown), by a driver according to a driving state of ones own vehicle. When ones own vehicle runs in an urban area, for instance, a driver operates the means for setting target to be searched in a back street so as to set a pedestrian as the target to be searched by the radar signal processor 12, and the means for setting target to be searched outputs a search signal, such as a pedestrian, to the means for determining forgetting factor FX, and the means for determining forgetting factor FX sets the small forgetting factor α the transient response of which is high in preparation for a pedestrian, a bicycle and a vehicle which suddenly rush in the forwarding direction.

If no big change in the state of a preceding vehicle in the forwarding direction is expected when ones own vehicle runs in a suburban area or on a highway, a driver sets the preceding vehicle as the target to be searched by the radar signal processor 12 with the operation of the means for setting target to be searched. Then, the means for setting target to be searched outputs a preceding vehicle searching signal to the means for determining forgetting factor FX, and the means for determining forgetting factor FX sets the big forgetting factor α for actualizing high S/N improvement in preparation for the search for the preceding vehicle in the forwarding direction. The forgetting factor can be thus properly changed according to kinds of the targets to be searched, such as a pedestrian and a preceding vehicle in the forwarding direction.

Another embodiment of the invention is now mentioned, referring to FIG. 23. The radar signal processor 12 of FIG. 23 has the observation means 1 through K denoted with reference numerals OB, means for respectively extracting distance components 1 through K denoted with reference numerals DS, the means for respectively estimating correlation matrices with exponential smoothing 1 through N denoted with reference numerals RM, the means for estimating presence and movement state of target denoted with the reference numeral OM, means for obtaining surroundings of ones own vehicle denoted with the reference numeral SN4, and the means for determining forgetting factor denoted with the reference numeral FX.

Explanations of the observation means p (array components P=1, 2, . . . K), and the means for extracting distance component p (array components P=1, 2, . . . K), the means for estimating correlation matrix with exponential smoothing q (distance rq: q=1, . . . , N), the means for estimating presence and movement state of target OM are omitted since these means are similar to ones of the embodiment of FIG. 1.

The means for obtaining surroundings of ones own vehicle SN4 obtains a speed limit in a road on which ones own vehicle runs, and outputs the obtained as information on speed limit AI4. As one of the methods of obtaining the surroundings of ones own vehicle, a peripheral image is obtained with a CCD sensor, and by picture processing and/or picture recognition on traffic control signs, the speed limit is known. Besides, whether the road on which ones own vehicle runs is a highway or an ordinary road may judged by the thus obtained speed limit information, and the judged may be outputted as information on road division.

The means for determining forgetting factor determines the forgetting factors α to be given to the means for estimating correlation matrix with exponential smoothing 1 through N on the basis of the information on speed limit AI4 or the information on road division. The optional forgetting factor α can be thus set according to the information on the speed limit or the road division.

The means for determining forgetting factor FX may determine the forgetting factors α to be given to the means for estimating correlation matrix with exponential smoothing 1 through N according to the speed limit obtained from the information on speed limit AI4. The speed limit is judged in such a way that the speed limit which is smaller than a predetermined threshold value is low speed limit, and the speed limit which is the same as or bigger than a predetermined threshold value is high speed limit. On the basis of such a judgment on speed limit, the small forgetting factor α of the most suitable predetermined ones is selected in case of high speed limit, and the big forgetting factor α of the most suitable predetermined ones is selected in case of low speed limit, and the selected is outputted as the forgetting factor α.

Besides, the means for determining forgetting factor FX may determine the forgetting factors α to be given to the means for estimating correlation matrix with exponential smoothing 1 through N according to the kinds of the road on which ones own vehicle runs which are obtained from the road division information, that is, a highway or an ordinary road. Whether the road is a highway or an ordinary road is judged in such a way that the speed limit which is smaller than a predetermined threshold value is an ordinary road, and the speed limit which is the same as or bigger than a predetermined threshold value is a highway. On the basis of such a judgment on kinds of roads, the small forgetting factor α of the most suitable predetermined ones is selected in case of a highway, and the big forgetting factor α of the most suitable predetermined ones is selected in case of an ordinary road, and the selected is outputted as the forgetting factor α.

By doing so, it is thus possible to suitably actualize high transient response performance of the correlation matrix estimation which is required for a target signal which is expected to change at the time of high speed limit, such as for a highway, and high S/N improvement performance which is required at the time of low speed limit, such as for an ordinary road.

Another embodiment of the invention is now mentioned, referring to FIG. 24. The radar signal processor 12 of FIG. 24 has a car navigation SN5 in place of the means for obtaining surroundings of ones own vehicle SN4 of the radar signal processor 12 of FIG. 23, so that explanations of the structural elements excluding the car navigation are omitted.

The car navigation SN5 gets the present position of ones own vehicle with a GPS, and outputs information on speed limit in the present position in map information which is stored in advance.

By doing so, effects similar to the invention of FIG. 23 can be exercised with the car navigation which can be easily mounted.

Another embodiment of the invention is now mentioned, referring to FIG. 25. The radar signal processor 12 of FIG. 25 has the observation means 1 through K denoted with reference numerals OB, means for respectively extracting distance components 1 through K denoted with reference numerals DS, the means for estimating correlation matrix with section average 1 through N denoted with reference numerals KS, the means for respectively estimating correlation matrices with exponential smoothing 1 through N denoted with reference numerals RM, the means for estimating presence and movement state of target denoted with the reference numeral OM, means for obtaining surroundings of ones own vehicle denoted with the reference numeral SN4, and the means for determining forgetting factor denoted with the reference numeral FX.

Besides, explanations of the means for obtaining surroundings of ones own vehicle SN4 and the means for determining forgetting factor FX are omitted since these means are similar to ones of the embodiment of FIG. 22.

Another embodiment of the invention is now mentioned, referring to FIG. 26. The radar signal processor 12 of FIG. 26 has the observation means 1 through K denoted with reference numerals OB, means for respectively extracting distance components 1 through K denoted with reference numerals DS, the means for respectively estimating correlation matrices with exponential smoothing 1 through N denoted with reference numerals RM, the means for estimating presence and movement state of target denoted with the reference numeral OM, means for obtaining radio wave state denoted with a reference numeral SN5, and the means for determining forgetting factor denoted with the reference numeral FX.

Explanations of the observation means p (array components P=1, 2, . . . K), and the means for extracting distance component p (array components P=1, 2, . . . K), the means for estimating correlation matrix with exponential smoothing q (distance rq: q=1, . . . , N) the means for estimating presence and movement state of target OM are omitted since these means are similar to ones of the embodiment of FIG. 1.

The means for obtaining radio wave state SN5 obtains information on radio wave state AI5 of the surroundings of ones own vehicle by analyzing waves which are received by a wave receiver. The information on radio wave state AI5 are a power of received wave, the presence of received wave and its power at the time of no transmitting of wave by the observation means. The obtained information on radio wave state AI5 is outputted to the means for determining forgetting factor FX.

The means for determining forgetting factor FX determines the forgetting factors α to be given to the means for estimating correlation matrix with exponential smoothing 1 through N on the basis of the information on radio wave state AI5. For instance, the small forgetting factor α of the most suitable predetermined ones is selected if the power of the received wave is higher than a predetermined threshold value, and the big forgetting factor α of the most suitable predetermined ones is selected if the power of the received wave is the same as or lower than a predetermined threshold value, and the selected is outputted as the forgetting factor α. Then, the forgetting factor α properly becomes bigger when the radio wave state is bad where an influence of noise is big due to small power of received wave, so that the high S/N improvement which is required at time of such a bad radio wave state can be realized.

The forgetting factor α can be thus optionally set according to the radio wave state.

Another embodiment of the invention is now mentioned, referring to FIG. 27. The radar signal processor 12 of FIG. 27 has the observation means 1 through K denoted with reference numerals OB, means for respectively extracting distance components 1 through K denoted with reference numerals DS, the means for respectively estimating correlation matrix with exponential smoothing 1 through N denoted with reference numerals RM, the means for estimating presence and movement state of target denoted with the reference numeral OM, means for obtaining radio wave state denoted with the reference numeral SN5, and the means for determining forgetting factor denoted with the reference numeral FX.

Explanations of the observation means p (array components P=1, 2, . . . K), and the means for extracting distance component p (array components P=1, 2, . . . K), the means for estimating correlation matrix with exponential smoothing q (distance rq: q=1, . . . , N), and the means for estimating presence and movement state of target OM are omitted since these means are similar to ones of the embodiment of FIG. 1.

The means for obtaining radio wave state SN5 obtains an interference influence AI6 by the other radar signal processors by analyzing waves which are received by a wave receiver. The interference influence AI6 is determined by a strength of the power of received wave at the time when the observation means transmits no wave so as to increase the interference influence according to the increase of the power. The obtained interference influence AI6 is outputted to the means for determining forgetting factor FX. “The interference by the other radar signal processors” is the interference by transmitted waves from the radar signal processor which is mounted on a vehicle, traveling in an opposite direction, for instance.

The means for determining forgetting factor FX determines the forgetting factors α to be given to the means for estimating correlation matrix with exponential smoothing 1 through N on the basis of the interference influence AI6. For instance, the small forgetting factor α of the most suitable predetermined ones is selected if the interference influence AI6 is higher than a predetermined threshold value, and the big forgetting factor α of the most suitable predetermined ones is selected if the interference influence AI6 is the same as or lower than a predetermined threshold value, and the selected is outputted as the forgetting factor α. Then, the influence by the past estimated value of the correlation matrix which has received the interference influence can be restricted since the forgetting factor α properly becomes smaller if the interference influence AI6 is big.

The forgetting factor α can be thus optionally set according to the interference influence.

At this time, zero (0) may be selected and outputted as the forgetting factor α if the interference influence is higher than a predetermined threshold value.

By doing so, the influence by the past estimated value of the correlation matrix which includes the interference influence can be completely removed since the forgetting factor α becomes zero (0) if the interference influence AI6 is very big.

Another embodiment of the invention is now mentioned, referring to FIG. 28. The radar signal processor 12 of FIG. 28 has the observation means 1 through K denoted with reference numerals OB, means for respectively extracting distance components 1 through K denoted with reference numerals DS, the means for estimating correlation matrix with section average 1 through N denoted with reference numerals KS, the means for respectively estimating correlation matrix with exponential smoothing 1 through N denoted with reference numerals RM, the means for estimating presence and movement state of target denoted with the reference numeral OM, means for obtaining radio wave state denoted with the reference numeral SN5, and the means for determining forgetting factor denoted with the reference numeral FX.

Besides, explanations of the means for obtaining radio wave state SN5 and the means for determining forgetting factor FX are omitted since these means are similar to ones of the embodiment of FIG. 26.

This invention can be utilized as the radar signal processor of a vehicle-mounted radar apparatus for detecting presence of preceding vehicles which exist in a forward direction by mounting on a vehicle.

The present invention has been explained on the basis of the example embodiments discussed. Although some variations have been mentioned, the embodiments which are described in the specification are illustrative and not limiting. The scope of the invention is designated by the accompanying claims and is not restricted by the descriptions of the specific embodiments. Accordingly, all the transformations and changes within the scope of the claims are to be construed as included in the scope of the present invention.

Claims

1. Radar signal processor, comprising: two or more observation means, each means for receiving a reflected wave from a target as a received signal through a receiving antenna and for outputting a predetermined observation signal; means for extracting distance component, for extracting a distance component corresponding to a distance from said receiving antenna from said each observation signal; means for estimating correlation matrix with exponential smoothing, for collecting respective said extracted distance components and for computing and estimating an estimated value of a correlation matrix corresponding to said respective distance components with a forgetting factor corresponding to said each distance component as a parameter by a method for estimating correlation matrix with exponential smoothing; means for estimating presence and movement state of target, for estimating a presence of a target or a movement state of said target on the basis of said respective estimated values of said correlation matrices which were estimated; and means for determining forgetting factor depending on distance, for respectively computing and determining said forgetting factors corresponding to said distances from said receiving antennas, and for respectively outputting said computed and determined forgetting factors as said forgetting factors corresponding to said respective distance components to said means for estimating correlation matrix with exponential smoothing.
2. Radar signal processor, comprising: two or more observation means, each means for receiving a reflected wave from a target as a received signal through a receiving antenna and for outputting a predetermined observation signal; means for extracting distance component, for extracting a distance component corresponding to a distance from said receiving antenna from said each observation signal; means for estimating correlation matrix with exponential smoothing, for collecting respective said extracted distance components and for computing and estimating an estimated value of a correlation matrix corresponding to said respective distance components with a forgetting factor corresponding to said each distance component as a parameter by a method for estimating correlation matrix with exponential smoothing; means for estimating presence and movement state of target, for estimating a presence of a target or a movement state of said target on the basis of said respective estimated values of said correlation matrices which were estimated; and means for determining forgetting factor depending on distance attenuation, for respectively computing and determining said forgetting factors corresponding to said distances from said receiving antennas according to an attenuation of said reflected wave due to said distance, and for respectively outputting said computed and determined forgetting factors as said forgetting factors corresponding to said respective distance components to said means for estimating correlation matrix with exponential smoothing.
3. Radar signal processor, comprising: two or more observation means, each means for receiving a reflected wave from a target as a received signal through a receiving antenna and for outputting a predetermined observation signal; range discrete Fourier Transform means, for respectively computing and extracting components of respective frequencies corresponding to said distance from said receiving antenna by discrete Fourier transform means on the basis of said observation signal; means for estimating correlation matrix with exponential smoothing, for collecting respective said extracted components of respective frequencies corresponding to said distances from said receiving antennas, and for computing and estimating an estimated value of a correlation matrix for said distance corresponding to said each frequency with a forgetting factor corresponding to said each distance as a parameter by a method for estimating correlation matrix with exponential smoothing; means for estimating presence and movement state of target, for estimating a presence of a target or a movement state of said target on the basis of said respective estimated values of said correlation matrices which were estimated; and means for determining forgetting factor depending on distance, for respectively computing and determining said forgetting factors corresponding to said distances from said receiving antennas, and for respectively outputting said computed and determined forgetting factors as said forgetting factors corresponding to said respective distance components to said means for estimating correlation matrix with exponential smoothing.
4. Radar signal processor, comprising: two or more observation means, each means for receiving a reflected wave from a target as a received signal through a receiving antenna and for outputting a predetermined observation signal; range discrete Fourier Transform means, for respectively computing and extracting components of respective frequencies corresponding to said distance from said receiving antenna by discrete Fourier transform means on the basis of said observation signal; means for estimating correlation matrix with exponential smoothing, for collecting respective said extracted components of respective frequencies corresponding to said distance from said receiving antenna, and for computing and estimating an estimated value of a correlation matrix for said distance corresponding to said each frequency with a forgetting factor corresponding to said each distance as a parameter by a method for estimating correlation matrix with exponential smoothing; means for estimating presence and movement state of target, for estimating a presence of a target or a movement state of said target on the basis of said respective estimated values of said correlation matrices which were estimated; and means for determining forgetting factor depending on distance attenuation, for respectively computing and determining said forgetting factors corresponding to said distances from said receiving antennas according to an attenuation of said reflected wave due to said distance, and for respectively outputting said computed and determined forgetting factors as said forgetting factors corresponding to said respective distance components to said means for estimating correlation matrix with exponential smoothing.
5. Radar signal processor, comprising: two or more observation means, each means for receiving a reflected wave from a target as a received signal through a receiving antenna and for outputting a predetermined observation signal; means for extracting distance component, for extracting a distance component corresponding to a distance from said receiving antenna from said each observation signal; means for estimating correlation matrix with section average, for collecting respective said extracted distance components and for computing and estimating a first estimated value of a correlation matrix corresponding to said respective distance components by a method for estimating correlation matrix with section average; means for estimating correlation matrix with exponential smoothing, for computing and estimating a second estimated value of a correlation matrix corresponding to said respective distance components with a forgetting factor corresponding to said each distance component as a parameter by a method for estimating correlation matrix with exponential smoothing from said first estimated value of said correlation matrix which was estimated; means for estimating presence and movement state of target, for estimating a presence of a target or a movement state of said target on the basis of said second estimated value of said correlation matrix which was estimated; and means for determining forgetting factor depending on distance, for respectively computing and determining said forgetting factors corresponding to said distances from said receiving antennas, and for respectively outputting said computed and determined forgetting factors as said forgetting factors corresponding to said respective distance components to said means for estimating correlation matrix with exponential smoothing.
6. Radar signal processor, comprising: two or more observation means, each means for receiving a reflected wave from a target as a received signal through a receiving antenna and for outputting a predetermined observation signal; means for extracting distance component, for extracting a distance component corresponding to a distance from said receiving antenna from said each observation signal; means for estimating correlation matrix with exponential smoothing, for collecting respective said extracted distance components and for computing and estimating an estimated value of a correlation matrix corresponding to said respective distance components with a forgetting factor as a parameter by a method for estimating correlation matrix with exponential smoothing; means for estimating presence and movement state of target, for estimating a presence of a target or a movement state of said target on the basis of said respective estimated values of said correlation matrices which were estimated; a ones own vehicle sensor, for sensing a movement state of ones own vehicle and for outputting the sensed as ones own vehicle movement state information; and means for determining forgetting factor, for respectively computing and determining said forgetting factors on the basis of said ones own vehicle movement state information and for outputting said computed and determined forgetting factors to said means for estimating correlation matrix with exponential smoothing.
7. The radar signal processor according to claim 6, wherein said ones own vehicle sensor senses a turning state of ones own vehicle, and outputs ones own vehicle turning state information which was thus obtained as said ones own vehicle movement state information.
8. The radar signal processor according to claim 6, wherein said ones own vehicle sensor has means for judging turning/non-turning, for judging whether ones own vehicle is turning or not, and for outputting a judged result as said ones own vehicle movement state information, and said means for determining forgetting factor has means for determining forgetting factor on turning state, for computing and determining said forgetting factor which has a low value if ones own vehicle is turning, and computing and determining said forgetting factor which has a high value if ones own vehicle is not turning on the basis of said ones own vehicle movement state information.
9. Radar signal processor, comprising: two or more observation means, each means for receiving a reflected wave from a target as a received signal through a receiving antenna and for outputting a predetermined observation signal; means for extracting distance component, for extracting a distance component corresponding to a distance from said receiving antenna from said each observation signal; means for estimating correlation matrix with exponential smoothing, for collecting respective said extracted distance components and for computing and estimating an estimated value of a correlation matrix corresponding to said respective distance components with a forgetting factor as a parameter by a method for estimating correlation matrix with exponential smoothing; means for estimating presence and movement state of target, for estimating a presence of a target or a movement state of said target on the basis of said respective estimated values of said correlation matrices which were estimated; a steering sensor, for detecting an actual angle of a steering of ones own vehicle and for outputting the detected as ones own vehicle turning state information; and means for determining forgetting factor, for respectively computing and determining said forgetting factors on the basis of said ones own vehicle turning state information, and for respectively outputting said computed and determined forgetting factors to said means for estimating correlation matrix with exponential smoothing.
10. Radar signal processor, comprising: two or more observation means, each means for receiving a reflected wave from a target as a received signal through a receiving antenna and for outputting a predetermined observation signal; means for extracting distance component, for extracting a distance component corresponding to a distance from said receiving antenna from said each observation signal; means for estimating correlation matrix with exponential smoothing, for collecting respective said extracted distance components and for computing and estimating an estimated value of a correlation matrix corresponding to said respective distance components with a forgetting factor as a parameter by a method for estimating correlation matrix with exponential smoothing; means for estimating presence and movement state of target, for estimating a presence of a target or a movement state of said target on the basis of said respective estimated values of said correlation matrices which were estimated; a Yaw Rate sensor, for measuring a turning angle of ones own vehicle and for outputting the measured as ones own vehicle turning state information; and means for determining forgetting factor, for respectively computing and determining said forgetting factors on the basis of said ones own vehicle turning state information, and for respectively outputting said computed and determined forgetting factors to said means for estimating correlation matrix with exponential smoothing.
11. Radar signal processor, comprising: two or more observation means, each means for receiving a reflected wave from a target as a received signal through a receiving antenna and for outputting a predetermined observation signal; means for extracting distance component, for extracting a distance component corresponding to a distance from said receiving antenna from said each observation signal; means for estimating correlation matrix with exponential smoothing, for collecting respective said extracted distance components and for computing and estimating an estimated value of a correlation matrix corresponding to said respective distance components with a forgetting factor as a parameter by a method for estimating correlation matrix with exponential smoothing; means for estimating presence and movement state of target, for estimating a presence of a target or a movement state of said target on the basis of said respective estimated values of said correlation matrices which were estimated; a vehicle speed sensor, for sensing a vehicle speed of ones own vehicle and for outputting the sensed as vehicle speed information; and means for determining forgetting factor, for respectively computing and determining said forgetting factors on the basis of said vehicle speed information, and for respectively outputting said computed and determined forgetting factors to said means for estimating correlation matrix with exponential smoothing.
12. The radar signal processor according to claim 11, wherein said vehicle speed sensor has means for judging vehicle speed, for judging whether ones own vehicle is running at a high speed or a low speed and for outputting the judged as said vehicle speed information, and said means for determining forgetting factor has means for determining forgetting factor on vehicle speed, for computing and determining said forgetting factor which has a small value if said vehicle speed which said vehicle speed information shows is high, and computing and determining said forgetting factor which has a high value if said vehicle speed which said vehicle speed information shows is low.
13. Radar signal processor, comprising: two or more observation means, each means for receiving a reflected wave from a target as a received signal through a receiving antenna and for outputting a predetermined observation signal; means for extracting distance component, for extracting a distance component corresponding to a distance from said receiving antenna from said each observation signal; means for estimating correlation matrix with section average, for collecting respective said extracted distance components and for computing and estimating a first estimated value of a correlation matrix corresponding to said respective distance components by a method for estimating correlation matrix with section average; means for estimating correlation matrix with exponential smoothing, for computing and estimating a second estimated value of a correlation matrix corresponding to said respective distance components with a forgetting factor as a parameter by a method for estimating correlation matrix with exponential smoothing from said first estimated value of said correlation matrix which was estimated; means for estimating presence and movement state of target, for estimating a presence of a target or a movement state of said target on the basis of said second estimated value of said correlation matrix which was estimated; a ones own vehicle sensor, for sensing a movement state of ones own vehicle and for outputting the sensed as ones own vehicle movement state information; and means for determining forgetting factor, for respectively computing and determining said forgetting factors on the basis of said ones own vehicle movement state information and for outputting said computed and determined forgetting factors to said means for estimating correlation matrix with exponential smoothing.
14. Radar signal processor, comprising: two or more observation means, each means for receiving a reflected wave from a target as a received signal through a receiving antenna and for outputting a predetermined observation signal; means for extracting distance component, for extracting a distance component corresponding to a distance from said receiving antenna from said each observation signal; means for estimating correlation matrix with exponential smoothing, for collecting respective said extracted distance components and for computing and estimating an estimated value of a correlation matrix corresponding to said respective distance components with a forgetting factor as a parameter by a method for estimating correlation matrix with exponential smoothing; means for estimating presence and movement state of target, for estimating a presence of a target or a movement state of said target on the basis of said respective estimated values of said correlation matrices which were estimated; means for obtaining surroundings of ones own vehicle, for obtaining surroundings of ones own vehicle and for outputting the obtained as information on surroundings of ones own vehicle; and means for determining forgetting factor, for respectively computing and determining said forgetting factors on the basis of said information on surroundings of ones own vehicle, and for respectively outputting said computed and determined forgetting factors to said means for estimating correlation matrix with exponential smoothing.
15. The radar signal processor according to claim 14, wherein said means for obtaining surroundings of ones own vehicle has means for obtaining information on speed limit, for obtaining a speed limit of a road where ones own vehicle is running and for outputting said obtained speed limit as said information on surroundings of ones own vehicle.
16. The radar signal processor according to claim 14, wherein said means for obtaining surroundings of ones own vehicle has means for judging speed limit, for judging as high speed limit if said speed limit of a road where ones own vehicle is running is higher than a predetermined value, and judging as low speed limit if said speed limit of said road where said ones own vehicle is running is lower than said predetermined value, and for outputting said judged high/low speed limit as said information on surroundings of ones own vehicle, and said means for determining forgetting factor has means for determining forgetting factor on speed limit, for computing and determining said forgetting factor which has a low value if said speed limit of said road where ones own vehicle is running is high, and computing and determining said forgetting factor which has a high value if said speed limit of said road where said ones own vehicle is running is low on the basis of said information on surroundings of ones own vehicle.
17. The radar signal processor according to claim 15, wherein said means for obtaining information on speed limit has car navigation means for getting a present position of ones own vehicle, and forgetting said speed limit of said present position from map information which is stored in advance.
18. Radar signal processor, comprising: two or more observation means, each means for receiving a reflected wave from a target as a received signal through a receiving antenna and for outputting a predetermined observation signal; means for extracting distance component, for extracting a distance component corresponding to a distance from said receiving antenna from said each observation signal; means for estimating correlation matrix with section average, for collecting respective said extracted distance components and for computing and estimating a first estimated value of a correlation matrix corresponding to said respective distance components by a method for estimating correlation matrix with section average; means for estimating correlation matrix with exponential smoothing, for computing and estimating a second estimated value of a correlation matrix corresponding to said respective distance components with a forgetting factor as a parameter by a method for estimating correlation matrix with exponential smoothing from said first estimated value of said correlation matrix which was estimated; means for estimating presence and movement state of target, for estimating a presence of a target or a movement state of said target on the basis of said second estimated value of said correlation matrix which was estimated; means for obtaining surroundings of ones own vehicle, for obtaining surroundings of ones own vehicle and for outputting the obtained as information on surroundings of ones own vehicle; and means for determining forgetting factor, for respectively computing and determining said forgetting factors on the basis of said information on surroundings of ones own vehicle and for outputting said computed and determined forgetting factors to said means for estimating correlation matrix with exponential smoothing.
19. Radar signal processor, comprising: two or more observation means, each means for receiving a reflected wave from a target as a received signal through a receiving antenna and for outputting a predetermined observation signal; means for extracting distance component, for extracting a distance component corresponding to a distance from said receiving antenna from said each observation signal; means for estimating correlation matrix with exponential smoothing, for collecting respective said extracted distance components and for computing and estimating an estimated value of a correlation matrix corresponding to said respective distance components with a forgetting factor as a parameter by a method for estimating correlation matrix with exponential smoothing; means for estimating presence and movement state of target, for estimating a presence of a target or a movement state of said target on the basis of said respective estimated values of said correlation matrices which were estimated; means for obtaining radio wave state, for obtaining radio wave state of surroundings of ones own vehicle from said received signal which was received by said receiving antenna, and for outputting the obtained as information on radio wave state; and means for determining forgetting factor, for respectively computing and determining said forgetting factors on the basis of said information on radio wave state, and for respectively outputting said computed and determined forgetting factors to said means for estimating correlation matrix with exponential smoothing.
20. The radar signal processor according to claim 19, wherein said means for obtaining radio wave state has means for computing interference, for computing and obtaining an interference influence due to waves excluding said signals transmitted from said observation means from said received signal, and for outputting said computed and obtained as said information on radio wave state.
21. The radar signal processor according to claim 20, wherein said means for determining forgetting factor has means for determining forgetting factor on interference influence, for computing and determining said forgetting factor as zero (0) if an interference influence in said information on radio wave state is higher than a predetermined threshold value.
22. Radar signal processor, comprising: two or more observation means, each means for receiving a reflected wave from a target as a received signal through a receiving antenna and for outputting a predetermined observation signal; means for extracting distance component, for extracting a distance component corresponding to a distance from said receiving antenna from said each observation signal; means for estimating correlation matrix with section average, for collecting respective said extracted distance components and for computing and estimating a first estimated value of a correlation matrix corresponding to said respective distance components by a method for estimating correlation matrix with section average; means for estimating correlation matrix with exponential smoothing, for computing and estimating a second estimated value of a correlation matrix corresponding to said respective distance components with a forgetting factor as a parameter by a method for estimating correlation matrix with exponential smoothing from said first estimated value of said correlation matrix which was estimated; means for estimating presence and movement state of target, for estimating a presence of a target or a movement state of said target on the basis of said second estimated value of said correlation matrix which was estimated; means for obtaining radio wave state, for obtaining a radio wave state of surroundings of ones own vehicle from said received signal received by said receiving antenna and for outputting the obtained as information on radio wave state; and means for determining forgetting factor, for respectively computing and determining said forgetting factors on the basis of said information on radio wave state and for outputting said computed and determined forgetting factors to said means for estimating correlation matrix with exponential smoothing.
23. Radar signal processor, comprising: two or more observation means, each means for receiving a reflected wave from a target as a received signal through a receiving antenna and for outputting a predetermined observation signal; means for estimating correlation matrix with exponential smoothing, for computing and estimating an estimated value of a correlation matrix corresponding to said observation signal on the basis of said each observation signal with a forgetting factor as a parameter by a method for estimating correlation matrix with exponential smoothing; means for estimating presence and movement state of target, for estimating a presence of a target or a movement state of said target on the basis of said respective estimated values of said correlation matrices which were estimated; a ones own vehicle sensor, for sensing a movement state of ones own vehicle and for outputting the sensed as ones own vehicle movement state information; and means for determining forgetting factor, for respectively computing and determining said forgetting factors on the basis of said ones own vehicle movement state information and for outputting said computed and determined forgetting factors to said means for estimating correlation matrix with exponential smoothing.
24. Radar signal processor, comprising: two or more observation means, each means for receiving a reflected wave from a target as a received signal through a receiving antenna and for outputting a predetermined observation signal; means for estimating correlation matrix with section average, for computing and estimating a first estimated value of a correlation matrix corresponding to said observation signal on the basis of said each observation signal by a method for estimating correlation matrix with section average; means for estimating correlation matrix with exponential smoothing, for computing and estimating a second estimated value of a correlation matrix from said first estimated value of said correlation matrix which was estimated with a forgetting factor as a parameter by a method for estimating correlation matrix with exponential smoothing; means for estimating presence and movement state of target, for estimating a presence of a target or a movement state of said target on the basis of said second estimated value of said correlation matrix which was estimated; a ones own vehicle sensor, for sensing a movement state of ones own vehicle and for outputting the sensed as ones own vehicle movement state information; and means for determining forgetting factor, for respectively computing and determining said forgetting factors on the basis of said ones own vehicle movement state information and for outputting said computed and determined forgetting factors to said means for estimating correlation matrix with exponential smoothing.
25. Radar signal processor, comprising: two or more observation means, each means for receiving a reflected wave from a target as a received signal through a receiving antenna and for outputting a predetermined observation signal; means for estimating correlation matrix with exponential smoothing, for computing and estimating an estimated value of a correlation matrix corresponding to said observation signal on the basis of said each observation signal with a forgetting factor as a parameter by a method for estimating correlation matrix with exponential smoothing; means for estimating presence and movement state of target, for estimating a presence of a target or a movement state of said target on the basis of said respective estimated values of said correlation matrices which were estimated; means for obtaining surroundings of ones own vehicle, for obtaining surroundings of ones own vehicle and for outputting the obtained as information on surroundings of ones own vehicle; and means for determining forgetting factor, for respectively computing and determining said forgetting factors on the basis of said information on surroundings of ones own vehicle and for outputting said computed and determined forgetting factors to said means for estimating correlation matrix with exponential smoothing.
26. Radar signal processor, comprising: two or more observation means, each means for receiving a reflected wave from a target as a received signal through a receiving antenna and for outputting a predetermined observation signal; means for estimating correlation matrix with section average, for computing and estimating a first estimated value of a correlation matrix corresponding to said observation signal on the basis of said each observation signal by a method for estimating correlation matrix with section average; means for estimating correlation matrix with exponential smoothing, for computing and estimating a second estimated value of a correlation matrix from said first estimated value of said correlation matrix which was estimated with a forgetting factor as a parameter by a method for estimating correlation matrix with exponential smoothing; means for estimating presence and movement state of target, for estimating a presence of a target or a movement state of said target on the basis of said second estimated value of said correlation matrix which was estimated; means for obtaining surroundings of ones own vehicle, for obtaining surroundings of ones own vehicle and for outputting the obtained as information on surroundings of ones own vehicle; and means for determining forgetting factor, for respectively computing and determining said forgetting factors on the basis of said information on surroundings of ones own vehicle and for outputting said computed and determined forgetting factors to said means for estimating correlation matrix with exponential smoothing.
27. Radar signal processor, comprising: two or more observation means, each means for receiving a reflected wave from a target as a received signal through a receiving antenna and for outputting a predetermined observation signal; means for estimating correlation matrix with exponential smoothing, for computing and estimating an estimated value of a correlation matrix corresponding to said observation signal on the basis of said each observation signal with a forgetting factor as a parameter by a method for estimating correlation matrix with exponential smoothing; means for estimating presence and movement state of target, for estimating a presence of a target or a movement state of said target on the basis of said respective estimated values of said correlation matrices which were estimated; means for obtaining radio wave state, for obtaining a radio wave state of surroundings of ones own vehicle from said received signal received from said receiving antenna and for outputting the obtained as information on radio wave state; and means for determining forgetting factor, for respectively computing and determining said forgetting factors on the basis of said information on radio wave state and for outputting said computed and determined forgetting factors to said means for estimating correlation matrix with exponential smoothing.
28. Radar signal processor, comprising: two or more observation means, each means for receiving a reflected wave from a target as a received signal through a receiving antenna and for outputting a predetermined observation signal; means for estimating correlation matrix with section average, for computing and estimating a first estimated value of a correlation matrix corresponding to said observation signal on the basis of said each observation signal by a method for estimating correlation matrix with section average; means for estimating correlation matrix with exponential smoothing, for computing and estimating a second estimated value of a correlation matrix from said first estimated value of said correlation matrix which was estimated with a forgetting factor as a parameter by a method for estimating correlation matrix with exponential smoothing; means for estimating presence and movement state of target, for estimating a presence of a target or a movement state of said target on the basis of said second estimated value of said correlation matrix which was estimated; means for obtaining radio wave state, for obtaining a radio wave state of surroundings of ones own vehicle from said received signal received from said receiving antenna and for outputting the obtained as information on radio wave state; and means for determining forgetting factor, for respectively computing and determining said forgetting factors on the basis of said information on radio wave state and for outputting said computed and determined forgetting factors to said means for estimating correlation matrix with exponential smoothing.
29. The radar signal processor according to claim 14, wherein said means for obtaining surroundings of ones own vehicle has means for judging information on road division, for judging whether a road where ones own vehicle is running is a highway or an ordinary road, and for outputting a judged result as information on surroundings of ones own vehicle.
30. The radar signal processor according to claim 29, wherein said means for determining forgetting factor has means for determining forgetting factor on road kind, for computing and determining said forgetting factor which has a low value if said road where ones own vehicle is running is a highway, and computing and determining said forgetting factor which has a high value if said road where ones own vehicle is running is an ordinary road on the basis of said information on surroundings of ones own vehicle.
31. Radar signal processor, comprising: two or more observation means, each means for receiving a reflected wave from a target as a received signal through a receiving antenna and for outputting a predetermined observation signal; means for extracting distance component, for extracting a distance component corresponding to a distance from said receiving antenna from said each observation signal; means for estimating correlation matrix with exponential smoothing, for collecting respective said extracted distance components and for computing and estimating an estimated value of a correlation matrix corresponding to said respective distance components with a forgetting factor as a parameter by a method for estimating correlation matrix with exponential smoothing; means for estimating presence and movement state of target, for estimating a presence of a target or a movement state of said target on the basis of said respective estimated values of said correlation matrices which were estimated; means for setting target to be searched, for selectively setting a kind of said target to be searched, and for outputting a corresponding search kind signal; and means for determining forgetting factor, for respectively computing and determining said forgetting factors on the basis of said search kind signal, and for outputting said computed and determined forgetting factors to said means for estimating correlation matrix with exponential smoothing.

Priority Claims (1)

Number	Date	Country	Kind
2005-100324	Mar 2005	JP	national

Radar signal processor

Information

Publication Number

Date Filed

Date Published

Inventors

Original Assignees

CPC

US Classifications

International Classifications

Abstract

Description

Claims

Priority Claims (1)