Patent Application
20060234699
The present invention relates to a system and a method for establishing the instantaneous speed in a reliable manner of an object which is travelling along a known trajectory, in particular a vehicle such as a train.
Patent application WO 02/03094 filed on 29 Jun. 2001 describes a system for locating an object in a reliable manner, such as a train, based on the transmissions of signals carried out by a group of GNSS (Global Navigation Satellite System) satellites.
In the context of railways, reliable location is understood in the sense of the standard Cenelec 50129. It relates to providing a location interval with an extremely low probability (from 10−09 to 10−12) that the train can be outside that interval.
With regard to establishing the instantaneous speed, prior systems conventionally based on electromechanical devices have the disadvantage of being extremely costly. In order to improve reliability, they use a combination of sensors which detect axle rotation and which have to withstand accelerations greater then 90 g, single-axle accelerometers and on-board radar systems.
Therefore, the object of the invention is to provide a system for establishing the instantaneous speed having the level of reliability required by railways at a cost far lower than existing systems.
Therefore, the invention relates to a system for establishing the instantaneous speed in a reliable manner of an object which is travelling along a known trajectory, in particular a vehicle such as a train, the system comprising:
Other features of the invention are:
The invention further relates to a method for establishing the instantaneous speed
of an object which is travelling along a known trajectory, which method comprises the steps of:
The invention will be better understood from a reading of the following description given purely by way of example and with reference to the drawings, in which:
The system for establishing the instantaneous speed is described with reference to
Let 1 designate an object which will be supposed to be a train, by way of example, which is travelling along a known trajectory 2.
It comprises location means 3 of the same type as those described in application PCT WO 02/03094 which are therefore capable of establishing, in a reliable manner, the position of the train 1 on the track based on signals from a group of GNSS satellites 4, 5 and maps of the network on which the train is travelling.
The train 1 is provided with means 6 for receiving signals transmitted by those satellites 4, 5. Those receiving means 6 comprise antennae and electronic modules which operate at superhigh frequency, as is well known to the person skilled in the art, and which are connected to the location means 3 in order to provide them with the signals from the satellites 4, 5.
The location means 3 are connected to a database 7 of maps of the rail network. That database conventionally depicts the tracks in the form of a succession of straight-line segments, each end of which is expressed using the WGS84 co-ordinate system of the GNSS system.
As explained in patent application WO 02/03094, those location means 3 provide a location interval of the train with a probability of less than 10−9 that the train can be outside that interval.
That location interval allows the movement direction to be established.
In accordance with the path of the track, the movement direction is corrupted by a variable angular imprecision. That imprecision is linked to two factors:
The signal received from the satellites 4, 5 by the receiving means 6 is also transferred to means 8 for calculating and analysing that signal.
They comprise (
Therefore, these measurement means 10 are connected to means 12 for establishing the vectorial difference of the speeds of the train and the satellite in the satellite/object direction by calculating the Doppler effect which has generated the frequency shift, which means 12 are themselves connected to means 13 for calculating the instantaneous speed in the direction of movement of the object based on the instantaneous speed established previously.
The frequency shift is generated by the Doppler effect according to the formula:
The means 12 for establishing the vectorial difference of the speeds comprise means 12A for calculating the speed {right arrow over (v)}s of the satellite using ephemerid data and the orbital model of the satellite, which data are downloaded from the satellites during a preceding step.
They also comprise means 12B for calculating the unit vector {right arrow over (a)} of the line of sight from the position of the object and the position of the satellite.
The means 12 for establishing the vectorial difference of the speeds of the train 1 and the satellite in the satellite/object direction, that is to say, ({right arrow over (v)}s−{right arrow over (v)}o)·{right arrow over (a,)}, calculate it in accordance with the equation:
which is deduced from equations (1) and (2).
The means 13 for calculating the instantaneous speed of the train in accordance with the movement direction thereof are brought about based on the instantaneous speed of the object in the satellite/object direction established previously, that is to say, {right arrow over (v)}o·{right arrow over (a)}, and the value of the vector {right arrow over (a)} expressed in the local reference system of the track (where the speed of the train has a unique component along the track).
These measurements and calculations are repeated for a plurality of different satellites independently. The combination of the results obtained carried out by means 14 then allows a confidence interval to be established for the speed with a low probability (from 10−09 to 10−12) of the speed being outside the interval.
In this manner, three measurements from three different satellites, each carried out with a confidence interval having an error probability of from 10−5 to 10−6, allow, at the first attempt, because they are independent measurements, a confidence interval to be obtained with a probability better than 10−15. In fact, it appeared that, owing to given errors, the probability obtained with three satellites is only in the order of 10−12.
The measurements carried out by the means 11 for measuring the frequency shift are corrupted by a given number of errors, such as the shift rate {dot over (δ)}to of the clock of the receiver relative to the time of the system, the time shift rate {dot over (δ)}tD owing to atmospheric phenomena, the noise of the receiver, . . . and the shift rate {dot over (δ)}ts of the clock of the satellite relative to the time of the system.
Should the train be provided with an atomic clock, as described above, the shift rate {dot over (δ)}to of the clock of the receiver may be considered to be negligible.
However, since an atomic clock is a relatively costly piece of equipment, it is particularly advantageous to have a system which does not require its use and which therefore allows the effect {dot over (δ)}to to be compensated for.
By expressing that frequency shift measurement in the form of a temporal variation of distance, that is to say, by multiplying the frequency shift owing to the Doppler effect by the wavelength of the signal, the temporal variation of the train/satellite distance measured
where {dot over (R)} is the temporal variation of the geometric train/satellite distance (in English “geometric range rate”).
Given that the shift rate of the clock of the satellite, derived from the navigation message of the satellite, is expressed as a constant referred to as af1,
then {dot over (ρ)}={dot over (R)}+c{dot over (δ)}tocaf1+c{dot over (δ)}tD (5).
By expressing the equation (3) in order to arrive at the temporal variation of distance,
By definition, that temporal variation of distance is equal to −λ·Doppler.
Thus, the temporal variation of distance measured {dot over (ρ)} is expressed as the negative product of the wavelength of the signal multiplied by the Doppler effect established by the receiver (L1Doppler), {dot over (ρ)}=−λT·L1 doppler.
Combining the equations (5) and (6) thus gives:
{right arrow over (v)}o·{right arrow over (a)}={right arrow over (v)}s·{right arrow over (a)}+λT·L1doppler+c{dot over (δ)}to−caf1+c{dot over (δ)}D (7).
Using the signals from two satellites s1 and s2, and combining the corresponding equations (7), there is obtained:
{right arrow over (v)}o·({right arrow over (a)}s2−{right arrow over (a)}s1)={right arrow over (v)}s2·{right arrow over (a)}s2−{right arrow over (v)}s1·{right arrow over (a)}s1+λT(L1Dopplers2−L1Dopplers1)−c(af1s2−af1s1)+c({dot over (δ)}tDS2−{dot over (δ)}tDS1).
The shift rate {dot over (δ)}to from the receiving clock is eliminated.
Thus, the means 12 for calculating the vectorial difference use the relative frequency shift owing to the Doppler effect of two signals from each of two separate satellites in order to eliminate the shift linked to the receiving clock, thus avoiding the use of a precision clock, such as an atomic clock, on the train 1.
In order to obtain the necessary precision for establishing the speed in a reliable manner, however, the measurements carried out with the signals of at least three satellites are necessary when an atomic clock is used, as explained above.
In order to obtain the same degree of precision without using an atomic clock and, therefore, using the calculations relating to two satellites explained above, it is necessary to use a group of at least four satellites, therefore forming six different pairs, in order to have the equivalent of three pairs of independent measurements.
The method for establishing the instantaneous speed of an object travelling along a known trajectory comprises the steps of:
All the above calculations have been carried out supposing that the signal follows a direct path between the satellite and the train.
It is well known that waves can be reflected from some surfaces and that it is therefore possible for the train to receive a reflected signal in place of the direct signal.
It will be appreciated that that leads to an additional level of error, falsifying the frequency shift of the signal received.
That “alternative path” phenomenon occurs only in a specific environment, such as travel through an urban zone.
A means of eliminating the signals arising from “alternative path” consists in installing two antennae at separate points of the train, typically approximately twenty metres from each other.
By carrying out a measurement of the Doppler effect on the signals received simultaneously at each antenna, it is established as to whether that measurement is different or not.
If it is different, that means that the signal received by one of the two antennae is from an “alternative path”.
In
A few moments later, the train having continued, the second antenna 21 enters the zone of the building 25 in turn.
At that moment, since the two antennae capture the same signal reflected by the building 25, the measurement means no longer detect any difference between the two signals.
Consequently, a time delay is provided in the form of a delay and a distance travelled, during which the signal from the satellite 4 is not used, and that time delay is re-initialised every time a divergence is observed. The satellite 4 in the example will therefore be re-used only after a given period of coherent measurement has been observed and/or the train has travelled a given distance without any divergence being observed. Other conditions may be associated therewith.
Therefore, the system and the method described in this manner advantageously allow the speed of a train to be established with the levels of reliability required. Therefore, it is far less costly than prior systems because it does not use any mechanical component or components subjected to powerful environmental constraints.
| Number | Date | Country | Kind |
|---|---|---|---|
| FR 0503803 | Apr 2005 | FR | national |
