GB2258965A - Doppler speed sensor - Google Patents

Abstract

A two-beam doppler speed sensor for use on a vehicle uses a single transmitter (11, 12) and receiver (13, 14). Two beams (21, 22) at 90 DEG are generated from a single beam (18) using quasi-optical components (23, 24, 25, 26). The two reflected beams (21', 22') are similarly combined using quasi-optics (29, 30, 31, 35) to give an output (16) from the receiver in which the two doppler frequencies of the beams are averaged spectrally. The beam-splitting components may be dielectric plates (24, 31) or polarisation-sensitive parallel conductor grids. In use, the sensor is aligned so that the two beams make equal angles of incidence (45 DEG ) with the ground, one beam looking forward, the other rearward. When through misalignment the incidence angles are not equal, one beam suffers a high doppler shift and the other beam suffers a low doppler shift, relative to the expected shift. The average peak spectral output from the receiver indicates the true velocity, the high and low doppler shifts effectively cancelling one another. Major advantage is a saving in components over designs which use two transmitters to generate the beams. <IMAGE>

Description

Doppler Speed Sensor This invention relates to doppler speed sensors for measuring a component of relative velocity between the sensor and a surface, which velocity component is parallel to the surface. Such sensors are commonly used to measure the speed of a vehicle relative to the ground. The angle of incidence which the emitted radio beam makes with the ground is a critical factor in the measurement accuracy which can be obtained. If this angle, typically 45 , differs from the intended value, through for example misalignment, an error is generated in the calculated speed.

One solution to this problem that has been adopted uses two transmitters to generate respective beams 90 apart, one directed forward the other rearward. If the two transmit antennas have a common mounting the angle between the beams is always maintained.

In this case, any error in the angles which the beams make with the ground will give rise to doppler outputs which are respectively high and low compared with the doppler frequency which is indicative of the true speed. However, when the two outputs are frequency averaged, the errors in the two outputs substantially cancel one another. Thus accuracy is improved over the single beam sensor by reducing the susceptibility to angle error.

Figure l(a) shows schematically the required hardware.

Each sensor comprises an oscillator 1, a transmit antenna 2, a receive antenna 3 and a mixer 4. The two transmit antennas 2 emit respective beams A and B which are separated by 90g. Reflections from the ground 7 are received by the receive antennas 3 and applied to one input of the mixers 4. The other input 5 to the mixers 4 is a sample of the output from the oscillator 1, which constitutes the "local oscillawor" signal. The mixers 4 thus each provide a homodyne output 6 which comprises the detected doppler frequency shifts. Figure l(b) shows the doppler frequency spectra of the two beams A and B when there is an angle error, i.e. the beams A and B make different angles of incidence with the ground. The two doppler outputs 6 are averaged in frequency by means of a summing amplifier 8.The spectrum of the output 9 of amplifier 8 is shown in Figure l(c). It gives a peak centred substantially on the "true" doppler frequency, this being equivalent to the relative velocity in the direction of the sensor beams. The necessary correction to derive the velocity component parallel to the ground is made by subsequent processing circuitry.

It is an object of the present invention to provide a doppler speed sensor which is capable of achieving an accuracy at least as good as the aforedescribed two-beam sensor, but which uses a reduced number of components.

According to the invention, there is provided a doppler speed sensor for measuring a component of relative velocity between the sensor and a surface, which velocity component is parallel to said surface, the sensor being adapted to provide two beams separated by a predetermined angle and directed in use towards the relatively oncoming part of said surface and the relatively receding part of said surface respectively, said two beams experiencing respective doppler frequency shifts from which said velocity component can be derived despite the two beams making different angles of incidence with said surface, wherein said two beams are generated from a single beam by beam splitting means. The invention thus provides a useful saving in components, the two beams being generated from the output of a single transmitter.

Preferably, the beam splitting means comprises quasi-optics.

In one embodiment the sensor includes an antenna for providing the single beam, the beam splitting means being adapted to combine the two reflected beams from said surface in a single receive beam incident on said antenna.

In a preferred embodiment there is a transmit antenna for providing the single beam, a receive antenna, and further beam splitting means adapted to combine the two reflected beams from said surface in a single receive beam incident on said receive antenna.

The or each beam splitting means may comprise a dielectric plate adapted to transmit and reflect respective portions of an incident beam, the two reflected beams being incident on opposite faces of one dielectric plate.

Preferably, the predetermined angle is 90 , the or each dielectric plate being disposed on and inclined at 45" to the axis of its associated antenna.

Alternatively, the or each beam splitting means may comprise a grid of parallel conductors arranged to transmit and reflect respective portions of an incident beam, the two reflected beams being incident on opposite faces of one grid.

Preferably, the predetermined angle is 90 , the or each parallel-conductor grid being disposed on and inclined at 45" to the axis of its associated antenna, with the conductors of the grid oriented at 450 to the E-plane of the associated single beam.

The or each beam splitting means preferably further comprises respective dielectric lenses for collimating the two beams and/or focusing the two reflected beams.

The or each antenna preferably comprises a planar array of microstrip patches.

Preferably, the sensor includes receiving means adapted to produce an output comprising the doppler frequency shifts, the sensor further comprising means responsive to the average peak frequency of the output to determine said velocity component.

One embodiment of the invention will now be described, by way of example only, with reference to the accompanying drawings, in which: Figure 1 shows a prior art sensor and associated doppler signal spectra, as referred to above; Figure 2 (a) shows schematically the hardware of the sensor to be described and Figure 2 (b) shows the doppler signal spectrum of the sensor output; and Figures 3 and 4 show respectively the arrangement of quasi-optical components used with the transmitter and the receiver of the sensor of Figure 2 (a).

Referring to Figures 2, 3 and 4, the sensor to be described comprises an oscillator 11 and mixer 14 (Figure 2 (a)). The sensor suitably operates in the millimetre wave radar band, for example at 94 GHz. A sample of the oscillator output is supplied to one input 15 of the mixer 14, the other input to the mixer 14 being supplied from a receive antenna 13 (Figure 4). The output of the oscillator 11 is supplied to a transmit antenna 12. Each antenna 12, 13 suitably comprises a planar array of microstrip patches (not shown) which produces a plane-polarised beam. Alternatively, horn antennas may be used.

The beam 18 emitted by the transmit antenna 12 is fed into an arrangement 20 of quasi-optical components which splits the beam into two beams 21, 22 separated by an angle of 90 (Figure 3). The arrangement 20 comprises a first dielectric lens 23 which focuses the beam 18, a dielectric plate 24, and two further lenses 25, 26 which respectively collimate the two output beams 21, 22 of the arrangement. The dielectric plate 24 is disposed on and inclined at 45 to the axis 27 of the transmit antenna 12. Lens 23 focuses the beam 18 on the dielectric plate 24. The plate 24 acts as a uO power splitter which splits the beam 18 by reflecting one portion of the beam 18 through 90' to produce beam 22, whilst transmitting another portion of the beam 18 to produce beam 21.These two half-power beams 21, 22 are collimated by the further lenses 25, 26 respectively to produce two narrow output beams of the sensor, which illuminate separate spots on the ground.

The reflections 21', 22' of the output beams 21, 22 are directed onto the receive antenna 13 by a similar arrangement 28 of quasi-optical components (Figure 4) as is used in conjunction with the transmit antenna 12. The reflected beams, separated by 90 and focused by lenses 29 and 30, are incident on opposite faces of a dielectric plate 31. The plate 31 acts in the same manner as plate 24 to split each incident beam into two half-power beams 32, 33, by reflecting one portion whilst transmitting another portion of each incident beam. The plate 31 is disposed on and inclined at 45e to the axis 34 of the receive antenna 13 so that beam 32 is focused by lens 35 onto the receive antenna 13. In this way the two reflected beams 21', 22' are combined in a single composite beam incident on the receive antenna 13. Although, as shown in Figure 4, half the reflected signal power is wasted (in beam 33), it will be appreciated that beam 33 could also be detected by the provision of a second receive antenna.

The output of the receive antenna 13 is supplied to one Input of the mixer 14 (Figure 2(a)), the other mixer input is being derived from the transmitter oscillator 11 by means for example of a coupler (not shown). The mixer 14 thus provides a homodyne output 16, i.e. having zero i.f. frequency, which contains the doppler frequency shifts to which the emitted beams 21, 22 are subject by virtue of the relative motion between the sensor and the ground 17.

In use, the sensor is aligned so that the two emitted beams 21, 22 have equal angles of incidence (45' in this case of two beams separated by 90') with the ground 17. Under these conditions, the two beams 21, 22 are directed towards relatively oncoming and relatively receding spots on the ground respectively. As a result, they will see relative velocities between the sensor and the ground which include velocity components along the ground in opposite directions. Since the two beams make the same angle with the ground, these velocity components will have equal magnitude. The reflected beams 21', 22' will thus reveal doppler frequency shifts of the same magnitude but of opposite sign, i.e. the reflected beams have frequencies which are respectively above and below the transmitted frequency by an amount dependent on the detected velocity component in the direction of the beams.The homodyne mixer output 16 does not distinguish between such 'positive' and 'negative' doppler frequencies, so that the spectral output will contain a single peak at the detected doppler frequency. The velocity component along the ground can be determined by subsequent processing circuitry (not shown) which takes account of the 45 angle between the actual measured velocity in the beam direction and the component along the ground which it is required to determine.

Consider now that the sensor has become misaligned so that the beams 21, 22 make different angles with the ground. The beams will now see relative velocities of different magnitude, giving rise to unequal doppler frequency shifts in the reflected beams, one below and one above the expected doppler frequency when the beams are both at 45 to the ground. The individual reflected beams 21', 22' will thus show spectral peaks at different frequencies.

However, once combined in a single beam 32 by the quasi-optical arrangement 28, the doppler spectrum of the mixer output will be inherently averaged. This spectrum is shown in Figure 2 (b), where It is assumed tlat the two doppler frequencies are sufficiently close that their spectral peaks overlap, as shown for example In Figure 1 (b), to give a 'flattened' overall peak. If the two doppler frequencies are further apart, the averaged spectrum will show two peaks. In all cases, however, the average peak doppler frequency in the mixer output will give a substantially corrected velocity reading, the higher doppler shift of one beam being compensated by the lower doppler shift of the other beam.Further processing circuitry (not shown) determines the peak average doppler frequency and makes the necessary calculation to determine the velocity component along the ground.

The described embodiment thus provides a two-beam sensor which requires only a single transmitter and receiver, plus the quasi-optics to generate the two output beams and to combine the reflected beams. There Is thus a significantly smaller number of hardware components than in the conventional two-beam sensor which requires a separate transmitter, receiver and antenna for each beam, in addition to a summing amplifier to average the two detected doppler signals. The quasi-optics also inherently generates two beams which are very accurately separated by 90', whereas the prior art sensor requires careful alignment of the transmit antennas to ensure the correct angle between the output beams. Furthermore, in the prior art sensor the two transmit oscillators must be accurately matched to avoid discrepancies between the two doppler frequencies.

This problem does not arise with the present Invention for which only a single transmit oscillator needs to be provided.

It will be appreciated that the output beams are not essentially separated by 90', provided the sensor is aligned initially so as that the two beams make equal angles of incidence with the ground. However, using a smaller angle reduces the useful velocity component along the ground. Further, beams separated by an angle of 90 can be conveniently generated. Thus 90 Is a preferred angle between the beams.

It will also be appreciated that it Is possible to use a single dielectric plate and one antenna for both transmission and reception, either by using a circulator or by adopting pulse transmission. The lenses used in conjunction with the plate or plates are not essential, but they are preferred since they assist in focusing and collimating the beams, which improves accuracy.

In an alternative embodiment of the invention (not illustrated), beam splitting is achieved by means of a grid comprising a plane array of parallel conductors at spacings comparable to the operational wavelength. It Is well-known that such a grid has the effect of transmitting a plane-polarised beam whose E-vector is at 90" to the conductors, whilst reflecting a plane-polarised beam whose E-vector is parallel to the conductors.

It will be apparent, therefore, that if the grid is oriented so that its conductors lie at 45e to the E-plane of a plane-polarised beam 18 from the transmit antenna 12, and the grid is inclined at 450 to the axis of the antenna 12, the incident beam will be part-reflected and part-transmitted by the grid to produce two beams 90' apart. A similarly arranged grid may be used to combine the two reflected beams. The dielectric plates 24, 31 in Figures 3 and 4 may thus be replaced by two such parallel-conductor grids. Implementing the beam splitting in this way may prove to be less expensive than the use of dielectric plates, since the grids are more easily fabricated.

It will be appreciated that since the grid is reciprocal In nature, a single grid and antenna may be used for both transmission and reception, again by using a circulator or by adopting pulse transmission.

Whereas operation at millimetre wave frequencies and the use of quasi-optics for the beam splitting is convenient and preferred, it will be appreciated that the principle of the invention applies to a much broader range of frequencies. Thus, for example, infrared frequencies may be used, in which case genuinely optical components will be needed for the beam splitting.

The invention is not limited to use on vehicles. The sensor may be mounted on the ground, for example at the roadside or trackside to detect the velocity of passing vehicles or trains.

Indeed, many other applications exist, in which the object whose speed is to be measured or monitored follows a substantially known path with respect to the sensor. One example is monitoring the speed of items carried on a conveyor.

Claims (12)

1. A doppler speed sensor for measuring a component of relative velocity between the sensor and a surface, which velocity component is parallel to said surface, the sensor being adapted to provide two beams separated by a predetermined angle and directed in use towards the relatively oncoming part of said surface and the relatively receding part of said surface respectively, said two beams experiencing respective doppler frequency shifts from which said velocity component can be derived despite the two beams making different angles of incidence with said surface, wherein said thv beams are generated from a single beam by beam splitting means.

2. A doppler speed sensor according to Claim 1, wherein said beam splitting means comprises quasi-optics.

3. A doppler speed sensor according to Claim 1 or Claim 2, including an antenna for providing said single beam, said beam splitting means being adapted to combine the two reflected beams from said surface in a single receive beam incident on said antenna.

4. A doppler speed sensor according to Claim 1 or Claim 2, including a transmit antenna for providing said single beam, a receive antenna, and further beam splitting means adapted to combine the two reflected beams from said surface in a single receive beam incident on said receive antenna.

5. A doppler speed sensor according to Claim 3 or Claim 4, as appendent to Claim 2, wherein the or each beam splitting means comprises a grid of parallel conductors arranged to transmit and reflect respective portions of an incident beam, said two reflected beams being incident on opposite faces of a said grid.

6. A doppler speed sensor according to Claim 5, wherein said predetermined angle Is 90', the or each parallel-conductor grid being disposed on and inclined at 45" to the axis of Its associated antenna, with the conductors of the grid oriented at 45 to the E-plane of the associated single beam.

7. A doppler speed sensor according to Claim 3 or Claim 4, wherein the or each beam splitting means comprises a dielectric plate adapted to transmit and reflect respective portions of an incident beam, said two reflected beams being incident on opposite faces of a said dielectric plate.

8. A doppler speed sensor according to Claim 7, wherein said predetermined angle is 90', the or each dielectric plate being disposed on and inclined at 45" to the axis of its associated antenna.

9. A doppler speed sensor according to any one of Claims 5 to 8, wherein the or each beam splitting means further comprises dielectric lenses for collimating said two beams and/or focusing said two reflected beams.

10. A doppler speed sensor according to any one of Claims 3 to 9, wherein the or each antenna comprises a planar array of microstrip patches.

11. A doppler speed sensor according to any one of the preceding claims, including receiving means adapted to produce an output comprising said doppler frequency shifts, the sensor further comprising means responsive to the average peak frequency of said output to determine said velocity component.

12. A doppler speed sensor substantially as hereinbefore described with reference to Figures 2, 3 and 4 of the accompanying drawings.