Patent Application
20110122013
The present invention relates to a radar apparatus that observes range and velocity of a vehicle by using an FMCW (Frequency Modulated Continuous Wave) system.
As a simple radar system for observing vehicles traveling on a road, an FMCW system is known (for example, refer to Non-patent Document 1). In the case of the radar apparatus employing the FMCW system, range and velocity of the vehicle are unknown quantities. Thus, in general, by combining up-chirp and down-chirp for transmission/reception waveform, the two parameters are calculated at the same time.
However, on the beat frequency axis, the frequencies of transmitted/received signals for the up-chirp and down-chirp are different even for the same target. Thus, when only a single target is present, the correspondence between the up-chirp and down-chirp can be established. However, when multiple targets are present, a problem occurs in that pairing the up-chirp and down-chirp for each target is difficult. Also, another problem occurs in that the cycle time tends to be long because the up-chirp and down-chirp need to be transmitted/received.
Also, when the frequency band related to the range resolution or the antenna aperture length related to the angular resolution is limited, still another problem occurs in that the resolution performance for dense targets is limited.
In addition to these problems, if an integral number N is small in the case of the same PRF (Pulse Repetition Frequency), each frequency bank width (PRF/N) on the beat frequency axis after FFT (Fast Fourier Transform) is performed on signals is increased, thus the frequency resolution deteriorates, and the accuracy of the range and velocity calculated based on the frequency is reduced.
Furthermore, when a complex signal is extracted by performing the Complex Fourier Transform on a real number signal to extract only a positive (or negative) frequency, if the actual sign of the target beat frequency is negative (or positive), correct range, velocity, and angle cannot be calculated.
The beat frequency becomes close to DC (frequency is 0) components for a short range. Still, the beat frequency needs to be separated from the DC even for a short range by increasing the frequency slope (frequency band B/sweep time T). In this case, if the frequency band B and the sample frequency PRF are limited, a problem occurs in that the integral number N cannot be made large. Especially, in the case of observing a target at a long range, smaller integral number causes reduced SN ratio (signal-noise ratio), thus the detection performance and accuracy are reduced.
A signal swept by a transmitter 21 inside the transmitter/receiver 20 is transmitted from an antenna transmission element 11. On the other hand, signals received by multiple antenna receive elements 12 each undergo frequency conversion by multiple mixers 22, then are sent to the signal processor 30. In the signal processor 30, the beat frequency signal from the transmitter/receiver 20 is converted to digital signals by an AD converter 31 to be sent to an up/down sequence extractor 37 as element signals (step S201).
The up/down sequence extractor 37 separates up-chirp and down-chirp signals from the element signals (digital signals) sent from the AD converter 31 to forward the up-chirp and down-chirp signals to an FFT unit 33 (step S202). The FFT unit 33 performs Fast Fourier Transform on the up-chirp and down-chirp signals sent from the up/down sequence extractor 37 to convert the signals into signals on the frequency axis, and forwards the signals to a DBF (Digital Beam Forming) unit 34.
The DBF unit 34 forms a Σ beam (up and down sequences) and a Δ beam by using the signals of the frequency axis sent from the FFT unit 33 (step S203). The Σ beam formed in the DBF unit 34 is sent to a pairing unit 38, and the Δ beam formed in the DBF unit 34 is forwarded to an angle measuring unit 36. The pairing unit 38 extracts frequencies of extreme amplitude as shown in
where
Δf1: observed frequency of down-chirp signal,
Δf2: observed frequency of up-chirp signal,
fd: Doppler frequency, and
fr: range-related frequency.
The range-related frequency fr and the Doppler frequency fd related to target velocity are given by the following equations.
Solving the equations (3) for the target range R, and target velocity V, and substituting the equations (2) into the resultant equations gives the following equations.
where
B: frequency band,
R: target range,
T: sweep time,
c: speed of light,
V: target velocity, and
λ: wave length.
When the above processing is completed, pairing of up sequence and down sequence is performed (step S205). That is, since the peak frequencies of down-chirp sequence and up-chirp sequence are different, processing to match a pair of frequencies is performed. The target range and velocity are then calculated (step S206), and the angle is calculated (step S207).
Subsequently, it is checked whether the cycle is completed or not (step S208). If the cycle is not completed in step S208, the process proceeds with processing of the next cycle (step S209). Subsequently, the process returns to step S201 and the above-described processing is repeated. On the other hand, if the cycle is completed in step S208, the processing by the radar apparatus is terminated.
By the above processing, the target range R and velocity V can be calculated. As described above, since the peak frequencies of down-chirp sequence and up-chirp sequence are different, a pair of frequencies needs to be matched. In the case of a single target or a small number of targets, the pairing is relatively easy. However, as the number of targets or reflection points in a background is increased, peak values on the frequency axis increase as shown
As described above, a conventional radar apparatus has the following problems.
(1) When up-chirp and down-chirp signals are combined as a transmission/reception waveform, if multiple targets are present, the pairing is difficult. Also, since the up-chirp and down-chirp signals need to be transmitted/received, the cycle time is increased.
(2) When the range resolution or angular resolution is limited, the resolution performance for dense targets is limited.
(3) If the integral number N is small in the case of the same PRF, each frequency bank width on the beat frequency axis after the Fast Fourier Transform is performed on signals is increased, thus the frequency resolution deteriorates, and the accuracy of the range and velocity calculated based on the frequency is reduced.
(4) When a complex signal is extracted by performing the Complex Fourier Transform on a real number signal to extract only a positive (or negative) frequency, if the actual sign of the target beat frequency is negative (or positive), correct range, velocity, and angle cannot be calculated.
(5) Since the beat frequency becomes close to DC (frequency is 0) components for a short range, the frequency slope needs to be increased. In this case, if the frequency band B and the sample frequency PRF are limited, the integral number N cannot be made large.
An object of the present invention is to provide a radar apparatus capable of observing a target with high detection performance and a high precision even if multiple targets are present in a wide area from a short range to a long range.
To solve the problems, the first invention includes: a transmitter/receiver that transmits an FMCW modulated sweep signal M times; an FFT unit that performs Fast Fourier Transform on the M sweep signals received in response to transmission from the transmitter/receiver; and an MRAV processor configured so that, when a maximum value of each sweep signal is calculated from the M sweep signals obtained by the Fourier Transform performed by the FFT unit, the MRAV processor performs: amplitude integration on beat frequency-sweep axis in a sweep direction for each of beat frequencies by using F (sweep number, target number) resulting from calculation of the beat frequencies by phase monopulse, amplitude monopulse, or MUSIC of M sweeps; calculates a least square line with respect to a relative range and a sweep time of a sweep number exceeding a predetermined threshold sweep for each frequency bank exceeding the predetermined threshold; calculates a target velocity from a slope of the least square line; and calculates a target range.
To solve the problems, the second invention includes: a transmitter/receiver that transmits an FMCW modulated sweep signal at least twice; an FFT unit that performs Fast Fourier Transform on the at least two sweep signals received in response to transmission from the transmitter/receiver; and an MRAV processor that calculates ranges and velocities of multiple targets by calculating beat frequencies corresponding to at least two sweeps by the transmitter/receiver based on the at least two sweep signals obtained by the Fourier Transform performed by the FFT unit, calculating velocities based on a frequency difference and a time difference of the calculated beat frequencies, and calculating ranges based on the calculated velocities and beat frequencies
To solve the problems, the third invention includes: a transmitter/receiver that transmits an FMCW modulated sweep signal M times; an FFT unit that perform Fast Fourier Transform on the M sweep signals received in response to the transmission from the transmitter/receiver; and an MRAV processor that performs smoothing over sweeps using F (sweep number, target number) resulting from calculation of beat frequencies by phase monopulse, amplitude monopulse, or MUSIC of the M sweeps when calculating a maximum value of each sweep signal from the M sweep signals obtained by Fourier Transform performed by the FFT unit, and calculates a range after calculating a velocity based on results of the smoothing.
To solve the problems, the fourth invention includes: a transmitter/receiver that transmits an FMCW modulated sweep signal M times; an FFT unit that performs Fast Fourier Transform on the M sweep signals received in response to transmission from the transmitter/receiver; and an MRAV processor that calculates a local maximum value on beat frequency-sweep axis by Hough transformation using F (sweep number, target number) resulting from calculation of beat frequencies by phase monopulse, amplitude monopulse, or MUSIC of the M sweeps when calculating a maximum value of each sweep signal from the M sweep signals obtained by Fourier Transform performed by the FFT unit, and calculates a range after calculating a velocity corresponding to the calculated local maximum value from a beat frequency difference and a sweep time
According to the first invention, on the beat frequency-sweep axis, by integrating the amplitude in the sweep direction for every beat frequency, an integration effect over multiple sweeps is obtained to improve the signal detection performance. Also, the range is calculated after the slope of a line extracted by fitting a least square line is calculated to determine the velocity. Accordingly, even if an error is present in the difference between the relative ranges, the influence of the error is reduced, and the accuracy in measuring the velocity and range can be improved.
According to the second invention, when multiple targets are present, pairing is not needed to be performed as in the conventional radar apparatus, and also, radar observation with a short cycle time may be achieved.
According to the third invention, the velocity and range are calculated by smoothing the relative range differences over sweeps, thus even if an error is present in the difference between the relative ranges, the influence of the error is reduced, and the accuracy in measuring the velocity and range can be improved.
According to the fourth invention, by performing the Hough transformation on the beat frequency-sweep axis, an integration effect over multiple sweeps is obtained to improve the signal detection performance. Also, the range is calculated after the slope of each line extracted by the Hough transformation is calculated to determine the velocity. Accordingly, even if an error is present in the relative range difference, the influence of the error is reduced, and the accuracy in measuring the velocity and range can be improved.
In the following, embodiments of the present invention are described in detail with reference to the drawings. The radar apparatus according to the present invention employs a simple system of pairing sequences between the banks with the same frequency or between neighboring banks by using FMCW signals having continuity, which are easy to be implemented.
A radar apparatus according to Embodiment 1 of the present invention employs MRAV (Measurement Range after measurement Velocity) system by which a range is measured after a velocity is measured by a beat frequency.
The antenna 10 is configured with an antenna transmitting element 11 and multiple antenna receiving elements 12. The antenna transmitting element 11 converts a transmission signal transmitted from the transmitter/receiver 20 as an electrical signal into a radio wave to send it to the outside. Multiple antenna receiving elements 12 receive radio waves from the outside to convert them into electrical signals, and send the signals as reception signals to the transmitter/receiver 20.
The transmitter/receiver 20 includes a transmitter 21 and multiple mixers 22. The multiple mixers 22 are provided for respective multiple antenna receiving elements 12. The transmitter 21 generates a transmission signal according to a transmission control signal sent from the signal processor 30, and sends the generated signal to the antenna transmitting element 11 and the multiple mixers 22. The multiple mixers 22 convert the frequencies of reception signals received from respective multiple antenna receiving elements 12 according to a signal from the transmitter 21, and forward the resultant signals to the signal processor 30.
The signal processor 30a includes an AD converter 31, an FFT unit 32, a DBF unit 34, an MRAV processor 35a, an angle measuring unit 36, and a transmission/reception controller 39.
The AD converter 31 converts an analog signal sent from the transmitter/receiver 20 into a digital signal according to a timing signal sent from the transmission/reception controller 39, and forwards the digital signal to the FFT unit 32 as an element signal.
The FFT unit 32 converts an element signal sent from the AD converter 31 into a signal on the frequency axis by the Fast Fourier Transform, and forwards the transformed signal to the DBF unit 34.
The DBF unit 34 forms Σ beam and Δ beam using the signal on the frequency axis sent from the FFT unit 33. The Σ beam formed in the DBF unit 34 is sent to the MRAV processor 35a, and the Δ beam formed in the DBF unit 34 is sent to the angle measuring unit 36.
The MRAV processor 35a measures range and velocity based on the Σ beam from the DBF unit 34. The range and velocity obtained by the range and velocity measurements in the MRAV processor 35a are outputted to the outside.
The angle measuring unit 36 measures an angle based on the Δ beam sent from the DBF unit 34. The angle obtained by angle measurement in the angle measuring unit 36 is outputted to the outside.
The transmission/reception controller 39 generates the transmission control signal to start transmission and sends the transmission control signal to the transmitter 21 of the transmitter/receiver 20 and also generates the timing signal to specify the timing at which a signal is taken in from the transmitter/receiver 20, and sends the timing signal to the AD converter 31.
Next, operations of the radar apparatus according to Embodiment 1 of the present invention configured as mentioned above are described with reference to the flowchart shown in
In measurement processing, when a cycle is started, first, the frequency changes continuously as shown in
In this state, Fast Fourier Transform (FFT) is performed (step S11). That is, the FFT unit 32 performs the Fast Fourier Transform on the element signal sent from the AD converter 31. Accordingly, as shown in
The DBF processing is then performed (step S12). That is, the DBF unit 34 forms Σ beam and Δ beam in the angular direction using the signal on the frequency axis sent from the FFT unit 33. Thus, as shown in
It is then determined whether the sweep is completed or not (step S13). That is, it is checked whether processing for both sweep 1 and sweep 2 is completed. In step S13, if the sweep is not completed, the process returns to step S11 and the processing described above is repeated for the sweep 2, which is the next FM modulated sweep signal.
On the other hand, if the sweep is completed in step S13, threshold level detection of the sweep 1 and sweep 2 is performed (step S14). That is, the DBF unit 34 detects a threshold level of the Σ beam obtained by the sweep 1 and sweep 2. Then the target detected in step S14 is stored (step S15). That is, the DBF unit 34 detects a target from the threshold level detected in step S14, and stores the target.
A beat frequency is then extracted (step S16). That is, the MRAV processor 35a extracts a beat frequency fp and a bank signal having a peak signal based on the result of performing the FFT and DBF on the sweep 1 and sweep 2 as shown in
The velocity V is then calculated (step S 17). That is, the MRAV processor 35a calculates relative ranges R1 and R2 using the beat frequency fp of the sweep 1 and sweep 2, and calculates the velocity V=(R2−R1)/T12.
The range R is then calculated (step S18).
where
R: range,
T: sweep time,
B: frequency band,
fp: beat frequency, and
c: speed of light.
The range R and velocity V are calculated from simultaneous equations based on the beat frequency fp and velocity V.
v: velocity,
R1, R2: ranges for the sweep 1 and 2,
T12: time interval between the sweep 1 and 2,
fp: beat frequency,
λ: wave length,
B: band, and
T: sweep time.
The angle θ is then calculated (step S19). That is, the angle measuring unit 36 measures angle based on the Δ beam sent from the DBF unit 34, and outputs the angle obtained by the angle measurement to the outside.
The target information is then stored (step S20). That is, the target velocity V calculated in the above-mentioned step S17, target range R, and target angle θ are stored. It is then checked whether the target is completed or not (step S21). That is, it is checked whether processing for all the targets is completed. In step S21, if the targets are not completed, current target number is changed to the next number and the process returns to step S16 to repeat the above-described processing. On the other hand, in step S21, if the targets are completed, the measurement processing is terminated.
As describe above, according to the radar apparatus according to Embodiment 1 of the present invention, the beat frequencies are the same because the signals of a down-chirp sequence or up-chirp sequence is transmitted/received, thus pairing does not need to be performed in the case of multiple targets. Also, radar observation with a short cycle time may be achieved.
Note that the radar apparatus according to Embodiment 1 described above first performs the Fast Fourier Transform (FFT), and then performs the DBF (Digital Beam Forming) to determine the beat frequency; however, as shown in
A radar apparatus according to Embodiment 2 of the present invention employs a system that combines phase monopulse with the MRAV system according to Embodiment 1 as described above. The configuration of the radar apparatus according to Embodiment 2 is the same as that of the radar apparatus according to Embodiment 1 shown in
Especially, in the case where the number of sample points is smaller for the same PRF, the interval between frequency banks is increased, and the frequency precision is reduced. Thus, as a measure against this problem, the radar apparatus according to Embodiment 2 uses phase monopulse used in the angle axis for the frequency axis as shown in
Monopulse measurement of the range and velocity calculates the error voltage εp of the following equation by using the Σ (f) and Δ (f) of extracted frequency of the target as shown in
Phase monopulse processing is performed by the FFT unit 32.
where
Σ: summation (after multiplying reception signals 1 to N by a weight 1, the FFT is performed on the signals),
Δ: subtraction (after multiplying reception signals 1 to N/2 by −1, and multiplying reception signals N/2+1 to N by a weight 1, the FFT is performed on the signals),
*: complex conjugate, and
Re: real part.
Then, the reference value ε0 of the error voltage εp calculated by using frequency characteristics of pre-stored Σ and Δ is arranged in a table (correspondence between ε0 and frequency f is made). The beat frequency fp is extracted from the above-mentioned observed value ε (step S16) by using the reference table. The velocity and range are then calculated using the extracted beat frequency fp (step S17, S18).
For the weight, in addition to −1 or 1, a weight such as a Taylor weight from a Taylor distribution may be used as a multiplier to reduce sidelobe. The Taylor distribution is described in, for example, “Takashi Yoshida (editorial supervision), ‘Radar Technology, revised version’, the Institute of Electronics, Information and Communication Engineers, pp. 274 and 275 (1996).”
As described above, according to the radar apparatus according to Embodiment 2 of the present invention, the beat frequency of each sweep signal is calculated with high precision based on the phase monopulse error voltage, thus the velocity and range may be calculated with high precision from a low velocity target to a high velocity target.
A radar apparatus according to Embodiment 3 of the present invention uses amplitude comparison monopulse instead of the phase monopulse of the radar apparatus according to Embodiment 2. The configuration of the radar apparatus according to Embodiment 3 is the same as that of the radar apparatus according to Embodiment 1 shown in
The Σ (f), Σ (f−1), and Σ (f+1) of the banks in the preceding and the following the extracted frequency of the target are used to compare the absolute values abs (Σ(f−1)) and abs (Σ(f+1)), and the larger one is set as abs (Σu).
Then the error voltage Ea of the following equation is calculated (see
where
Σ: summation (after multiplying reception signals 1 to N by a weight 1, the FFT is performed on the signals), and
Σu: either (f−1) or E(f+1) that has larger absolute value.
Then, the reference value ε0 of the error voltage εp calculated by using frequency characteristics of pre-stored absolute values abs (Σ) and abs (Σu) is arranged in a table (correspondence between ε0 and frequency f is made). The beat frequency fp is extracted based on the above-mentioned observed value 68 by using the reference table. The velocity and range are calculated by using the extracted beat frequency fp.
As described above, according to the radar apparatus according to Embodiment 3 of the present invention, the beat frequency of each sweep signal is calculated with high precision based on the amplitude monopulse error voltage, thus the velocity and range may be calculated with high precision from a low velocity target to a high velocity target.
For the weight, similarly to the above-described radar apparatus according to Embodiment 2, in addition to −1 or 1, a weight such as a Taylor weight may be used as a multiplier to reduce sidelobe.
A radar apparatus according to Embodiment 4 of the present invention uses MUSIC system. The MUSIC system is described in “HARRY B. LEE, “Resolution Threshold of Beamspace MUSIC For Two Closely Spaced Emitters”, IEEE Trans. ASSP, Vol. 38, No. 9, Sept. (1990).”
As shown in
The above-mentioned document describing the MUSIC system describes the beam in terms of the angle axis, however the beam may be extended to a frequency axis such as that obtained by Fast Fourier Transform (FFT).
The process is as follows. That is, as shown in
R
bb
=X·X
H [Equation 8]
where
Rbb: correlation matrix,
X: column vector having (2M+1) elements of X-M to X0 to XM, and
H: complex conjugate transposition.
The eigenvectors Eb of the correlation matrix Rbb are then calculated. And an eigenvector EN with respect to noise is extracted from the eigenvectors Eb. By using the eigenvector EN and a steering vector w to search for the frequency, MUSIC spectrum is calculated by the following equation. As shown in
Smusic: MUSIC spectrum, and
w: steering vector of the frequency axis.
where
ws: steering column vector having elements ws(n) on the time axis,
fs: search frequency,
PRF: repeat frequency,
n: 1 to N (N is a sample number),
FFT[ ]: Fourier Transform,
EN: eigenvector of the correlation matrix Rbb, and
H: conjugate transposition.
The velocity and range are then calculated by using the determined beat frequency fp. (steps S17, S18).
As described above, according to the radar apparatus according to Embodiment 4 of the present invention, the beat frequency of each sweep signal is calculated with high precision based on the FFT and the MUSIC processing, thus the velocity and range may be calculated with high precision from a low velocity target to a high velocity target.
As shown in
This radar apparatus transmits an FMCW modulated sweep signal N times (#1 to #N), and extracts local maximum values at Nt points from the result of the Fast Fourier Transform for each sweep. As shown in
First, N sweeps are transmitted/received and the first FFT is performed on each sweep by the FFT unit 32 (steps S11 to S13). A target bank that exceeds a predetermined threshold is then extracted (step S51). The target bank of each sweep then undergoes the second FFT by the second FFT unit 39 (step S52). Subsequently, the beat frequency fp of a peak is read. And by using the calculated beat frequency fp, calculation of the velocity (step S17) and calculation of the range (step S18) are performed.
In the case of phase monopulse, processing is performed as follows. First, N sweeps are transmitted/received and the first FFT is performed on each sweep by the FFT unit 32 (steps S11 to S13). The Σ signal and Δ signal are then calculated, and a target bank whose absolute value of the Σ signal exceeds a predetermined threshold is extracted (step S51). The Σ signal and Δ signal of the target bank of each sweep then undergo the second FFT by the second FFT unit 40 (step S52). The frequency fp is then calculated by using the Σ signal and Δ signal, and the range and velocity are calculated by using the calculated frequency fp.
As described above, according to the radar apparatus according to Embodiment 5 of the present invention, by further performing the second FFT over sweeps on the extracted bank signal of each sweep, the target positions and velocities can be extracted with even higher resolution than in the bank obtained by the first FFT only.
A radar apparatus according to Embodiment 6 of the present invention calculates, when respective local maximum values are calculated from two sets of M sweep signals in the radar apparatus according to Embodiment 5, the Σ and Δ in the M sweep signals as shown in
First, N sweeps are transmitted/received and the first FFT is performed on each sweep (steps S11 to S13). A target bank whose amplitude exceeds a predetermined threshold is extracted (step S51). The Σ and Δ are calculated by the second FFT 40 unit using the target banks E1 to EM extracted in step S51. Then, the error voltage εp of the following equation is calculated by using the Σ (f) and Δ (f) of extracted frequency of the target.
Σ: summation (after multiplying E1 to EM by a weight 1, the FFT is performed thereon),
Δ: subtraction (after multiplying E1 to EM/2 by a −1, and multiplying EM/2+1 to EM by a weight 1, the FFT is performed thereon),
*: complex conjugate, and
Re: real part.
Then, the reference value ε0 of the error voltage εp calculated by using frequency characteristics of pre-stored Σ and Δ is arranged in a table (correspondence between ε0 and frequency f is made). The frequency value fp is extracted from the above-mentioned observed value ε (step S16) by using the reference table. And by using the calculated beat frequency fp, calculation of the velocity (step S17) and calculation of the range (step S18) are performed.
As described above, according to the radar apparatus according to Embodiment 6 of the present invention, by calculating the frequency with the phase monopulse as the second FFT is performed by the second FFT unit 40, the beat frequency can be extracted with high precision and the target position and velocity can be extracted with high precision.
For the weight, similarly to the above-described radar apparatus according to Embodiment 2, in addition to −1 or 1, a weight such as a Taylor weight may be used as a multiplier to reduce sidelobe.
A radar apparatus according to Embodiment 7 of the present invention, when respective local maximum values are calculated from M sweep signals in the radar apparatus according to Embodiment 6, calculates the Σ and Σu in the M sweep signals as shown in
First, N sweeps are transmitted/received and the first FFT is performed on each sweep (steps S11 to S13). A target bank whose amplitude exceeds a predetermined threshold is extracted (step S51). Then the Σ (f), Σ (f−1), and Σ (f+1) of the banks in the preceding and the following the extracted frequency of the target are used to compare the absolute values abs (Σ (f−1)) and abs (Σ (f+1)), and the larger one is set as abs (Σu).
And the error voltage εa of the following equation is calculated. This processing for the monopulse error voltage is performed by the second FFT unit 40 (step S61).
where
Σ: summation (after multiplying reception signals 1 to N by a weight 1, the FFT is performed on the signals), and
Σu: either Σ (f−1) or Σ (f+1) that has larger absolute value.
Then, the reference value ε0 of the error voltage εp calculated by using frequency characteristics of pre-stored absolute values abs (Σ) and abs (Σu) is arranged in a table (correspondence between ε0 and frequency f is made). The beat frequency fp is extracted based on the above-mentioned observed value ε by using the reference table. The velocity and range are calculated by using the extracted beat frequency fp.
As described above, according to the radar apparatus according to Embodiment 7 of the present invention, by calculating the frequency with the amplitude monopulse as the second FFT is performed by the second FFT unit 40, the beat frequency can be extracted with high precision and the target position and velocity can be extracted with high precision.
For the weight, similarly to the above-described radar apparatus according to Embodiment 2, in addition to −1 or 1, a weight such as a Taylor weight may be used as a multiplier to reduce sidelobe.
A radar apparatus according to Embodiment 8 of the present invention, when respective local maximum values are calculated from two sets of M sweep signals, calculates the beat frequency by performing the FFT and the MUSIC processing on the M sweep.
First, based on the result of the FFT in steps S11 to S13, a target bank whose amplitude exceeds a predetermined threshold is extracted (step S51). A correlation matrix Rbb is then calculated based on the complex signals Xm for M sweeps of the target bank.
R
bb
=X·X
H [Equation 12]
where
Rbb: correlation matrix,
X: column vector having (M) elements of X1 to XM, and
H: complex conjugate transposition.
The eigenvectors Eb of the correlation matrix Rbb are then calculated. And an eigenvector EN with respect to noise is extracted from the eigenvectors Eb. By using the eigenvector EN and a steering vector w to search for the frequency, MUSIC spectrum is calculated by the following equation, and the beat frequency fp at which the spectrum has an extremum is read. The MUSIC processing is performed by the second FFT unit 39 (step S71).
Smusic: MUSIC spectrum, and
w: steering vector of the frequency axis
where
ws: steering column vector having elements ws(n) on the time axis,
fs: search frequency,
Fs: 1/sweep time interval,
n: 1 to N (N is a sample number),
FFT[ ]: Fourier Transform,
EN: eigenvector of the correlation matrix Rbb, and
H: conjugate transposition.
Then by using the calculated beat frequency fp, calculation of the velocity (step S17) and calculation of the range (step S18) are performed.
As described above, according to the radar apparatus according to Embodiment 8 of the present invention, by calculating the frequency with the FFT and the MUSIC processing as the second FFT is performed by the second FFT unit 40, the beat frequency can be extracted with high precision and the target position and velocity can be extracted with high precision.
A radar apparatus according to Embodiment 9 of the present invention uses a system in which when the sweep signal is a real number signal (not a complex number signal), Complex Fourier Transform is performed on a sampled signal to extract a positive (or negative) signal from the beat frequencies. In this case, even if actual target has a negative (or positive) beat frequency, the beat frequency may be observed as a positive (or negative) frequency depending on the range and velocity of the target, thus the range and velocity may be miscalculated. In this case, it can be determined whether the beat frequency is negative (or positive) by the following principle.
According to the MRAV system, a velocity moving away is observed as v′=−v (R2<R1)<0 (see
fp′=−2v/λ+2R·B/cT>0.
fp′+2v′/λ=2R·B/cT
From the third one of Equation (6), R′=(positive term)·(fp′+2v′/λ)=(positive term)×2R·B/cT<0.
Therefore, the range R′ of the above equation is negative, and in this case, the velocity is determined to be the one moving away. If the velocity is determined to be negative, correct values of v, R may be obtained by reversing respective signs of v′, fp′, and R′.
A negative velocity observation system according to the MRAV system is executed by the following process. After the difference between the ranges in two sweeps is observed, the target velocity V is calculated based on positional change and time (step S17). Then by using the target velocity V calculated in step S17, the range R is calculated from the beat frequency (step S18). The angle θ is then calculated (step S19). It is then checked whether the range R is negative (step S61). If the range R is negative in step S61, respective signs of the range R, velocity V, and angle θ are reversed by a sign reversing unit (not shown) (step S62). On the other hand, if the range R is not negative in step S61, processing in step S62 is skipped.
As described above, according to the radar apparatus of Embodiment 9 of the present invention, in the case where Complex Fourier Transform is performed on a real number sampling frequency to observe only positive (or negative) beat frequency and yet actual target signal has negative (or positive) beat frequency, the range, velocity, and angle can be converted so as to have correct signs after determining the correct signs by the sign of calculated range.
A radar apparatus according to Embodiment 10 of the present invention uses the system employed by Embodiment 1 or Embodiment 2 in the case where observation targets are present in a wide area from a short range to a long range, and velocity range is wide.
In the case of a short range target, a transmission signal becomes close to the DC on the beat frequency axis after the FFT, and tends to be easily influenced by noise due to a sneaking of the transmission signal to the receiving side. On the other hand, in the case of a long range target, frequencies are more separated from the DC components, thus the influence of noise tends to be smaller. Thus, the transmission/reception signals are divided into for a short range and for a long range as shown in
When the band B and the PRF are the same, for a short range, the number of sample points is small; however, noise reduction effect is greater than the integral effect of the signals, thus desired SN ratio can be secured. For a long range, the number of sample points is large, thus the integral effect of the signals is great, and desired SN ratio can be secured.
As described above, the radar apparatus according to Embodiment 10 of the present invention is capable of transmitting/receiving a signal having a large integral number according to the range from a short range to a long range with a low noise frequency separated from the DC components on the beat frequency axis, thus radar observation with a high SN ratio can be achieved.
By setting a shorter time between sweeps for the case of faster target velocity, and a longer time between sweeps for the case of slower target velocity, the accuracy in velocity measurement can be improved. Thus, when the target velocities are unknown, by discriminating those targets with a high critical factor from others, for which higher velocity accuracy is desired, an optimal sweep needs to be selected focused on the targets.
As a weight indicator of critical factor, heavier weight may be placed on a target with a high relative velocity, approaching in a short range, and may be expressed by the following equation.
Cr=k×V/R (14)
where
Cr: critical factor,
k: constant,
V: relative velocity, and
R: relative range.
In order to maximize the accuracy of observation, the sweep Ts (or a sweep number close to the Ts) may be selected by the following equation, assuming that the target velocity is V and the frequency bank width is Af.
If only the absolute values of Cr are used for the selection, a positive velocity (the velocity for approaching), and a negative velocity (the velocity for moving away) are mixed in the selection. Since a positive velocity having higher critical factor cannot have higher priority in the case where the absolute value of a negative velocity is greater, processing is performed by the above-described process.
As described above, according to the radar apparatus according to Embodiment 11 of the present invention, velocity accuracy can be improved with a shorter time between sweeps for the case of faster target velocity, and a longer time between sweeps for the case of slower target velocity, thus when the target velocity is unknown, an optimal sweep according to the target may be selected by determining a target for which improved velocity accuracy is desired using its critical factor.
In the case of multiple targets, by periodically changing to a different sweep for every cycle rather than determining a critical factor, accuracy of the velocity and range may be improved for every target. A radar apparatus according to Embodiment 12 of the present invention periodically changes to a different sweep for every cycle.
First, a sweep set is selected (step S81). Then an initial sweep is set (step S82). A sweep is then set (step S83). It is then checked whether the cycle is completed or not (step S84). In step S84, if the cycle is not completed, the sweep is changed (step S85). Subsequently, the process returns to step S83 to repeat the processing described above. On the other hand, if the cycle is completed in step S84, the processing is terminated, and the process proceeds with the processing of the next cycle.
As described above, according to the radar apparatus according to Embodiment 12 of the present invention, accuracy of the velocity and range may be improved for every target by periodically changing to a different sweep for every cycle.
Since the MRAV system described above depends on the accuracy of a relative range difference (beat frequency) created from two sweeps, if the SN ratio is low, the velocity and range accuracy are reduced. In order to improve this situation, a radar apparatus according to Embodiment 13 of the present invention uses M sets of multiple sweep signals to obtain a smoothing effect as shown in
Smoothing is performed by a smoothing filter over S (m, n) sweeps (step S95), and after the velocity is calculated by using this result, the relative range is calculated (steps S17 and S18). The processing in steps S91 to S94, step S95, and steps S17 and S18 is performed by the MRAV processor 35.
Rs(1, n)=R(1, n) Equations 16
Rs(m, n)=R(m)+α(R(m)−Rs(m−1)) (15)
where
Rs: smooth range,
m: sweep number (m=1 to M),
n: target number, and
α: constant.
The range R and velocity V are calculated from simultaneous equations based on fp and V.
v: velocity,
T1M: time interval between the sweeps 1 and M,
fp: beat frequency,
λ: wave length,
B: band, and
T: sweep time.
The smoothing filter is only used to obtain a smoothing effect, and other filter such as a Least Square filter may also be used. Although the first and the Mth sweeps are used for calculating the velocity, other sweep interval may also be used as long as a smoothing effect can be obtained.
As described above, according to the radar apparatus according to Embodiment 13 of the present invention, a relative range difference for calculating the velocity is observed by multiple sweeps (multiple time intervals) and is smoothed. Accordingly, even if an error is present in the relative range difference, the influence of the error is reduced, and the accuracy in measuring the velocity and range can be improved.
The MRAV system described above depends on the accuracy of a relative range difference (beat frequency) from two sweeps. Thus, if the SN ratio is low, the velocity and range accuracy are reduced. In order to improve this situation, a radar apparatus according to Embodiment 14 of the present invention uses M sets of multiple sweep signals and performs the Hough transformation to obtain an integration effect as well as a smoothing effect as shown in
And F (m, n) is used to perform the Hough transformation by the beat frequency-sweep axis (step S101) so that local maximum values (ρp, θp) (p=1 to P) exceeding a predetermined threshold is extracted (step S102). Then a line is calculated by the following relational equation for every local maximum value, and the beat frequency fp is calculated from the intersection between this line and the beat frequency axis X.
ρp=X cos θp+Y sin θp (17)
where
X: beat frequency, and
Y: sweep number.
The beat frequency fp is then converted into a relative range using Equations (5), and the velocity Vp is calculated by the following equation (step S17).
Vp=(RM−R1)/(TM−T1)
where
R1, RM: relative ranges on the X-axis (beat frequency axis) from the lines corresponding to the sweep 1 and sweep M, and
T1, TM: times of the starting points of the sweep 1 and the sweep M.
Subsequently, by using the beat frequency fp and velocity Vp, the range is calculated from the following equation (step S18). The processing of steps S92 to S94, steps S101 and 102, and steps S17 and S18 are performed by the MRAV processor 35.
Vp: velocity,
fp: beat frequency,
λ: wave length,
B: band, and
T: sweep time.
Although the beat frequency axis-sweep axis has been used when the Hough transformation is performed, the Hough transformation may be used on the relative range-sweep axis after the beat frequency is converted to a relative range using Equations (5).
Here, general Hough transformation is described. The Hough transformation is the technique of extracting a line from an image. A line on the X-Y plane expressed in the polar coordinate has following equation as shown in
ρ=X cos θ+Y sin θ
By the above equation, the line, ρ, and θ uniquely correspond. Next, as shown in
(1) A matrix to store numerical values on the ρ-θ axis is reserved.
(2) Centered on an observed value on the X-Y axis, ρ on the ρ-θ axis is calculated for θ sequentially changed by Δθ and 1 is added to the element at the corresponding line, column of the matrix. This processing is repeated for all of the observed values.
(3) A local maximum point (ρq, θq) (q=1 to Q) is extracted from the matrix.
By the above steps, Q lines can be extracted from (ρq, θq).
As described above, according to the radar apparatus according to Embodiment 15 of the present invention, by observing relative range differences for calculating velocities by multiple sweeps (multiple times) to perform the Hough transformation on the relative range difference-sweep time axis, an integration effect over multiple sweeps is obtained to improve the signal detection performance. Also, by calculating the slope of each line extracted by the Hough transformation to determine the velocity, then later the range, even if an error is present in the relative range difference, the influence of the error is reduced, and the accuracy in measuring the velocity and range can be improved.
Since the MRAV system described above depends on the accuracy of a relative range difference (beat frequency) from two sweeps, if the SN ratio is low, the velocity and range accuracy are reduced. In order to improve this situation, a radar apparatus according to Embodiment 15 of the present invention uses M sets of multiple sweep signals and performs amplitude integration to obtain an integration effect as well as a smoothing effect as shown in
And F (m, n) is used to perform amplitude integration (video integration) for every frequency bank (step S111) on the beat frequency-sweep axis so that frequency banks fb exceeding a predetermined threshold is extracted (step S112).
Subsequently, for the frequency bank fb, using F (m, n) for the sweep exceeding the threshold, the least square line with respect to the sweep time m and the relative range Rp is calculated (step S114), which may be expressed as below, then the velocity is calculated from the slope of the least square line (step S17).
Rp=a·t+b
where
Rp: relative range,
t: starting time of sweep,
a: line slope (corresponding to the velocity Vp), and
b: constant.
Subsequently, by using the beat frequency fp and velocity Vp, the range is calculated from the following equation (step S18). The processing of steps S92 to S94, steps S111 to 114, and steps S17 and S18 are performed by the MRAV processor 35.
Vp: velocity,
fp: beat frequency,
λ: wave length,
B: band, and
T: sweep time.
For fp used in Equations (19), an average value in each sweep may be used.
As described above, according to the radar apparatus according to Embodiment 15 of the present invention, by observing relative range differences for calculating velocities by multiple sweeps (multiple times) to perform the video integration on the relative range difference-sweep time axis, an integration effect over multiple sweeps is obtained to be able to improve the signal detection performance. By calculating the slope of the line extracted by fitting the least square line, the range is calculated, then later the velocity is calculated. Thereby, even if an error is present in the relative range difference, the influence of the error is reduced, and the accuracy in measuring the velocity and range can be improved.
The present invention may be used for a radar apparatus that measures the range to a vehicle and the velocity of the vehicle.
| Number | Date | Country | Kind |
|---|---|---|---|
| 2009-121902 | May 2009 | JP | national |
| Filing Document | Filing Date | Country | Kind | 371c Date |
|---|---|---|---|---|
| PCT/JP2010/054839 | 3/19/2010 | WO | 00 | 12/3/2010 |
