Polarimetric radar for object classification and suitable method and suitable use therefor

Abstract

The present invention relates to a polarimetric radar, consisting of a transmission assembly that emits circularly polarized waves by means of transmission antennas and a receiver assembly that receives the reflected circularly polarized wave components by means of an antenna assembly. A plurality of two-channel receivers are provided as the receiver assembly, which simultaneously receive clockwise-rotating and anti-clockwise-rotating circularly polarized signal components, which are provided for digital beam shaping downstream of the antenna assembly. The invention further relates to a method for object classification.

Claims

1. A polarimetric radar, comprising: a transmission assembly comprising transmission antennas that emits circularly polarized waves during operation of the polarimetric radar; and a receiver assembly that receives reflected components of the circularly polarized wave using an antenna assembly during operation of the polarimetric radar, wherein the receiver assembly comprises a plurality of two-channel receivers which, during operation, simultaneously receive clockwise- and anticlockwise-rotating circularly polarized signal components with a common phase center which are provided for digital beam shaping downstream of the antenna assembly.

2. The polarimetric radar according to claim 1, wherein the transmission assembly comprises a plurality of transmitters in a horizontal direction and a vertical direction, the plurality of transmitters having phase center distances which are chosen dependently upon the phase center spacing of individual antenna elements of the antenna assembly of the receiver assembly such that periodically recurring main lobes are suppressed.

3. The polarimetric radar according to claim 2, wherein for digital beam steering in the vertical and the horizontal direction each individual antenna element or groups of antenna elements of the antenna assembly of the receiver assembly has a separate receiving channel for co-polar received signals relating to the transmitted signal and for cross-polar received signals relating to the transmitted signal and half-line transmission location changeover is realized in both the vertical and in the horizontal direction by the use of at least 4 transmission antennas.

4. The polarimetric radar according to claim 2, wherein the receiver assembly has vertical receiving lines forming a receiving network designed such that the vertical receiving lines, which comprise a real component and a synthetic component generated by half-line transmission location changeover, have a low level of sub-lobes by means of amplitude superposition and amplitude multiplication factors of the synthetic lines.

5. The polarimetric radar according to claim 1, wherein each transmission and reception antenna decouples right- and left-circularly polarized waves.

6. The polarimetric radar according to claim 1, wherein the distance between two phase centers of individual antennas elements of the receiver assembly has a value between the wavelength and 1.25 times the wavelength of the carrier frequency of an irradiated wave of the radar system.

7. The polarimetric radar according to claim 1, wherein the transmission antennas have an aperture size compatible to half-line transmission location changeover, the lobe width of which covers the scanning range of the radar receiver during the transmission process.

8. The polarimetric radar according to claim 1 wherein the transmission assembly consists of 6 transmission antennas.

9. The polarimetric radar according to claim 1, wherein each of the transmitters emits during operation, reversibly, both a left- and a right-circularly polarized wave.

10. A method for object classification using a polarimetric radar according to claim 1, comprising: a) providing a transmission assembly that emits circularly polarized electromagnetic waves and a plurality of two-channel receivers as the receiver assembly which receive reflected electromagnetic waves by means of an antenna assembly using digital beam shaping, both the clockwise-rotating and the anticlockwise-rotating circularly polarized wave components being received simultaneously and with a common phase center, and b) classifying the objects according to type and size by the position of reflection focal points of the objects, the reflection focal points being determined by range and speed Fourier transforms both in the receiving channel for anticlockwise- and in the receiving channel for clockwise-rotating circularly polarized received signals by means of frequency modulation.

11. The method according to claim 10, wherein, in order to determine the position of the reflection focal points of the target objects of the co-polar received signals relating to the transmitted signal, and in order to determine the position of the reflection focal points of the target objects of the cross-polar received signals relating to the transmitted signal, a range and speed Fourier transform is respectively calculated and a spectrum is evaluated for the object classification.

12. The method according to claim 10, wherein, in order to determine the position of the reflection focal points of the target objects of the co-polar received signals relating to the transmitted signal with respect to the position of the reflection focal points of the target objects of the cross-polar received signals relating to the transmitted signal, a common range and speed Fourier transform is calculated and the spectrum for the object classification is evaluated.

13. The method according to claim 10, wherein in order to determine the position of the reflection focal points of the target objects of the co-polar received signals relating to the transmitted signal with respect to the position of the reflection focal points of the target objects of the cross-polar received signals relating to the transmitted signal, a range and speed Fourier transform is calculated by means of one of the two receiving channels in order to determine an approximate range gate of the relevant objects and then a high-resolution discrete range and speed Fourier transform is calculated for both receiving channels separately and for both receiving channels together by means of the respective range gate with the relevant objects and their spectra are evaluated for the object classification.

14. A method, comprising: using the polarimetric radar according to claim 1 for object determination while integrated into a moving base.

15. The polarimetric radar of claim 5, wherein the antenna assembly comprises an axially constructed corrugated horn and the right- and left-circularly polarized waves are decoupled with a common phase center and with an integral septum polarizer when using the axially constructed corrugated horn.

16. The polarimetric radar of claim 7, wherein the half-line transmission location changeover has a lobe width of 3 dB.

17. The polarimetric radar of claim 8, wherein the transmission assembly comprises two adjacent horizontally arranged transmission antennas which are vertically arranged in triplicate, adjacent transmission antennas having spacing which is (n?0.5) times a spacing of the phase center of the individual receiver antennas and n is a whole number.

18. The polarimetric radar of claim 9, wherein the antenna assembly comprises an axially constructed corrugated horn and the right- and left-circularly polarized waves are decoupled with a common phase center and with an integral septum polarizer when using the axially constructed corrugated horn.

19. The method of claim 14, wherein the moving base is a car.

20. The method of claim 14, wherein the polarimetric radar is used with a transmission frequency permitted for automotive applications.

21. The method of claim 14, wherein the polarimetric radar is used with a transmission frequency in a frequency range of 76 GHz to 81 GHz.

22. The method of claim 14, wherein the polarimetric radar is used by emitting a frequency modulated continuous wave signal.

Description

DETAILED DESCRIPTION

(1) The polarimetric radar with digital beam shaping for object classification uses circular polarisation with the carrier frequency of 76 to 81 GHz permitted for automotive applications. Both the clockwise-rotating and the anticlockwise-rotating circular signal components which are reflected on the object are evaluated. Receiver-side digital beam shaping is used for the geometric angular resolution of the received data. This principle is applied in order to be able to calculate the complete radar image from one measurement for a fixed time. The disadvantage that the size of the individual radar backscatter cross-sections as e.g. in mechanically or electronically scanning systems is changed during the scan is thus avoided. This is a crucial advantage in order to be able to undertake a reliable polarimetric evaluation of the radar backscatter cross-sections of the object.

(2) FIG. 1 shows the arrangement of the transmission and reception antenna units. In the horizontal direction the receiving unit consists of a number of receiving channels (Rx). Each receiving channel consists of individual emitters which are connected to one another vertically by a feed network. This arrangement enables digital beam shaping in the horizontal direction. By this principle the received data of the individual receiving channels which are detected at the same time are multiplied by an amplitude assignment in the subsequent digital signal processing, a phase component is applied to them and they are then added up so that the direction of the main lobe of the overall array can be swivelled. FIG. 2 shows the dependency between the swivel angle and phase components which can be added to the individual phases of the receiving channels. These are dependent, moreover, upon the distance between the receivers and the wavelength of the carrier frequency.

(3) The permissible frequency range for automotive applications is between 76 GHz and 81 GHz, i.e. the wavelength is approx. 4 mm. In order to avoid grating lobes, according to antenna theory the horizontal distance between receiving antennas should be half the wavelength of approx. 2 mm. In practice however it is not possible to arrange the receiving antennas so closely together. This gives rise to over-coupling, and the required isolation between the co-polars and cross-polars of at least 20 dB in order to be able to carry out the polarimetric target classification is lost. According to the application, by means of a greater distance between the receiving antennas, the creation of grating lobes by semi-line transmission location changeover can be avoided. With half-line transmission location changeover one transmits alternately with 2 spatially offset transmitters and one receives with the identical receiver array. The spatial offset of the transmitters is chosen here such that the receiving antennas thus lie virtually and centrally between the receiving antennas of the real receiving array. By adding the measurements with the first and the second transmitter when shaping the digital beam, the condition of the half wavelength between the reception antennas is fulfilled once again, and the creation of grating lobes is prevented.

(4) As well as the receiving array, FIG. 1 also shows an arrangement of 6 transmission antennas. By using a number of transmitters which are spaced apart by a distance ds equal to (n?0.5)*dz, ds standing for the vertical and horizontal distances between the phase centres of the transmitters, and dz standing for the vertical and horizontal distances between the phase centres of the individual emitters of the receiver, and n being a natural smallest possible number, half-line transmission location changeover can be implemented by temporal transmission location changeover. FIG. 3 shows, as an example, this principle for 2 transmitters. The transmitters are used alternately, and the distance between the transmitters is imaged onto the receiving array so that a synthetic receiving array is produced that is shifted by the distance between the transmitter and the real receiving array. By means of the aforementioned correlation between the transmitter distances according to the formula, the synthetic receivers lie between the real receivers and one obtains halving of the receiver distances which should have approximately half the wavelength of the carrier frequency of the radar system in order to avoid disruptive grating lobes (periodically recurring main lobes) and to realise large swivel angles by means of digital beam shaping. If the parameter n becomes greater than 1 here, no distance halving takes place at the edge of the receiving array. Since these gaps at the edge of the antenna array cause secondary lobes, the corresponding channels are not taken into account for the digital beam shaping.

(5) By means of half-line transmission location changeover the distances between the phase centres of the real reception antennas is doubled with the same performance. The space gained in this way is advantageous for the technical producibility of complex antennas. When forming the overall array the change to the phase when changing signal run times must be corrected by object movements relative to the radar sensor during the transmission process. So that half-line transmission location changeover can also be used for objects with smaller and average ranges, the transmitting and receiving unit must be arranged close to one another.

(6) The multiple use of semi-line transmission location changeover in the horizontal and the vertical direction for the arrangement from FIG. 1 is shown by FIG. 4. For reasons relating to space the 6 transmitters have phase sensor spacing here which is 1.5 times the spacing of the phase centres of the individual reception antennas. In the receiving array gaps are therefore created on the edge of the array where no halving of the spacing of the receiving lines takes place. According to the above description the outer synthetic receiving line and the outer real receiving line are therefore not taken into account in the digital beam shaping. FIG. 5 shows the transfer of the transmission and reception assembly with 12 real receiving lines in one layout that shows the space required by the corrugated horn antennas used with circles. The distances for the transmitters and receivers are chosen here such that one can cover an angle range of approximately +/?45? by means of digital beam shaping. The number of e.g. 12 receiving lines leads to a 3 dB beam width of approx. 8?.

(7) The feed network for a real receiving line which is constructed symmetrically must be designed here as regards hardware such that the assigned amplitude of the whole receiving line, that consists of the real and synthetic individual receiving antennas, guarantees high sub-lobe suppression. The synthetic individual reception antennas are produced by the real receiving line being shifted upwards and downwards by 1.5 times the spacing of the phase centres of the individual reception antennas by means of half-line transmission location changeover with 3 transmitters. Here the synthetic individual reception antennas are superposed in specific positions so that the corresponding assigned amplitude coefficients and multiplication factors are produced for the synthetic receiving lines which are shown in FIG. 6.

(8) The reception antennas must receive the clockwise-rotating and the anticlockwise-rotating circularly polarised signal components simultaneously here. According to FIG. 7 two receiving channels are produced for one receiving line for both polarisation directions. The requirement here is for high decoupling between co- and cross-polar signal components. This is achieved by the corrugated horn emitter with a septum polarizer shown in FIG. 8. According to FIG. 5 the space requirement for a receiver is 4.2 mm. This value is approximately the wavelength of the carrier frequency of 76 to 77 GHz which is approx. 3.9 mm. The diameter of the aperture opening is designed to be 4.1 mm. According to the invention the septum polarizer has the property of a common phase centre for the receiving channel for the anticlockwise- and for the receiving channel for the clockwise-rotating circularly polarised received signals. This is a crucial pre-requisite for the polarimetric object classification.

(9) During the transmission process only an anticlockwise-rotating or a clockwise-rotating or a temporally alternately clockwise- or anticlockwise-rotating circularly polarised wave is emitted (FIG. 9). Here a corrugated horn emitter is once again suitable which, according to FIG. 10, has a space requirement of 6.3 mm and is designed for an aperture opening diameter of 6.1 mm. The size of the aperture opening must be chosen such that the main lobe of the transmitter covers the scanning range of the radar.

(10) At the target object the polarisation is changed according to the surface structure of the object. Here larger target objects can be deconstructed into a number of individual targets, as e.g. shown in a greatly simplified manner in a vehicle in FIG. 11. The change in polarisation direction is determined here by the following properties of the object: angle of symmetry, orientation angle to the irradiating wave, number of reflections in the object and angle of polarisability. The latter is a measure for how strongly an object can polarise an unpolarised electromagnetic wave. As shown in FIG. 11, the properties of the reflection focal point decides whether, for example, a circularly anticlockwise-rotating wave at a reflection focal point on the object, that for example, shown in simplified form, only contains total reflections, is reflected as a circularly anticlockwise-rotating or circularly clockwise-rotating wave and is thus received. Since a circular wave changes the direction of rotation by total reflection, the following situation arises: reflection focal points of the object with an even number of total reflections do not bring about any change to the direction of rotation between the received and the transmitted signal, e.g. dihedral reflection focal points with an odd number of total reflections bring about a change to the direction of rotation between the received and the transmitted signal, e.g. corner reflector.

(11) By means of the geometric representation of the object from reflection focal points with an even and an odd number of total reflections it is possible, for example, to determine the physical dimensions of the object and so the type or object category.

(12) Another key aspect of the invention is the algorithm for determining the reflection focal points of the target. For this purpose the transmitting signal is frequency-modulated (FMCW) according to FIG. 13. According to the FMCW principle known from the literature, one then obtains in the received signal the so-called range gate according to the level of the frequency deviation of the transmitted signal and after calculation of the Fourier transform (range Fourier transform) with the following range resolution:
range resolution=light speed/(2 times frequency deviation)
and after another calculation of a fast Fourier transform (FFT) over a number of frequency ramps (Doppler FFT) the speed information for the objects in the individual range gates. According to FIG. 12 the range and speed FFT for the receiving channel are calculated here in parallel for the receiving channel for anticlockwise- and clockwise-rotating received signals, i.e. the reflection focal points with an even number of total reflections appear in the spectrum of the channel without polarisation rotation, and those with an uneven number of total reflections appear in the channel with polarisation rotation. From the spectrum the distance between equal polarisation focal points can then respectively be determined. In order to additionally be able to determine the distance between the rotating and non-rotating polarisation focal points, a common FFT of double the length is calculated over both channels. Here e.g. the values of the non-rotating channel are set on the real part of the input sequence of the FFT, and the values of the rotating channel are set on the imaginary part. The object is classified by evaluating the spectral components of the 3 spectra calculated by means of FFT.

(13) Instead of calculating respectively a high-resolution FFT (anticlockwise-rotating and clockwise-rotating and sum channel), the following algorithm is also possible: The range and the speed FFT are calculated with low range resolution (small frequency deviation). If a possible object has then been identified in a range gate on the basis of the backscatter cross-section, for this specific range gate a high-resolution DFT (discrete Fourier transform) is then calculated in the rotating and in the non-rotating and in the sum channel by a new transmitting/receiving cycle taking place with a high frequency deviation. The object classification takes place as specified above.

DESCRIPTION OF THE FIGURES

(14) FIG. 1 shows the arrangement of the transmission and reception antenna unit. This consists of 6 transmitters which have spacing that is (n?0.5) times the spacing dz of the phase centre of the individual reception antennas, n being a natural number. 3 vertically arranged sensors are respectively located here at two horizontal positions. The receiving array is formed from a number of horizontally arranged receiving channels which consist of individual reception antennas in the vertical direction. The distances from the phase centres of all of the individual reception antennas correspond to the parameter dz.

(15) The symbols indicate as follows here:

(16) Rx: receiving channel

(17) Tx: transmitter

(18) ds: phase centretransmitter spacing

(19) dz: phase centreindividual reception antenna spacing

(20) n: natural number (1, 2, 3, . . . )

(21) Y: individual element antenna

(22) FIG. 2 shows the principle of digital beam shaping. Here, an additional phase component is applied to each receiver. The amount of this phase component is dependent upon the swivel angle, the geometric arrangement of the receiver, the distance between the receivers and the wavelength of the carrier frequency of the radar system.

(23) The symbols indicate as follows here:

(24) Y: receiver

(25) ?: phase components

(26) d: distance between two receivers

(27) ?: swivel angle

(28) ?.sub.0: wavelength

(29) FIG. 3 shows the mode of operation of half-line transmission location changeover. Here two transmitters are arranged with spacing (n?0.5) times the spacing dz of the phase centre of the real individual receiving antennas, n being a natural number. On the receiver side a receiving array is thus created consisting of real and synthetic receivers, the latter being located between the real receivers, with the exception of the array edge.

(30) The symbols indicate as follows here:

(31) Rx: real receivers

(32) Rx_s: synthetic receiver

(33) Tx: transmitter

(34) dz: phase centreindividual reception antenna spacing

(35) n: natural number (1, 2, 3, . . . )

(36) FIG. 4 shows the overall receiving array consisting of the real and the synthetic component and the associated 6 transmitters which having spacing 1.5 times the spacing dz of the phase centre of the real individual reception antennas. Here each real and synthetic individual antenna is shown symbolically, and the part of the receiving array used in the signal processing is encircled.

(37) The symbols indicate as follows here:

(38) : synthetic individual antenna

(39) Y: real individual antenna

(40) Tx: transmitter

(41) Rx: receiving line

(42) dz: phase centreindividual reception antenna spacing

(43) FIG. 5 shows the transmitting and receiving unit from FIG. 4, the receiving array consisting of 12 receiving channels, and only the real individual antennas being shown, and indeed the space required by the latter when using corrugated horn antennas. The spacing of the phase centres of the individual antennas is specified as 4.2 mm. The resulting precise spacings for the transmitting and receiving unit are given in brackets in millimeters.

(44) The symbols indicate as follows here:

(45) O: space requirement for an individual antenna (corrugated horn antenna)

(46) dz: phase centreindividual reception antenna spacing

(47) Tx: transmitter

(48) Rx: receiving line

(49) FIG. 6 shows the vertical assigned amplitude of a receiving line when using 3 vertical transmitters according to FIG. 5. Shown here are the space requirement of the real individual receiving antennas of a receiving line, the respectively shifted synthetic individual receiving antennas, the respective assigned amplitude coefficients of each individual receiving antenna, and additionally the multiplication factor for the synthetic antennas. The assigned amplitude of the real individual reception antennas is symmetrical here. The receiving line (on the left) consists of real and synthetic horn emitters. The transmitting unit consists of three vertical transmitters.

(50) The symbols indicate as follows here:

(51) Tx: transmitter

(52) Rx: receiving line

(53) dz: phase centreindividual reception antenna spacing

(54) a: assigned amplitude coefficient of the individual reception antenna

(55) s: multiplication factor of a synthetic receiving line

(56) O: space requirement for a real individual antenna (corrugated horn antenna)

(57) ?: shifted synthetic individual receiving antenna

(58) FIG. 7 shows the receiving channel structure of a receiving line. For each individual antenna there is a receiving channel for right-circularly polarised signals and a receiving channel for left-circularly polarised signals. All of the receiving channels for right-circularly and for left-circularly polarised signals are connected to one another here.

(59) The symbols indicate as follows here:

(60) RHC=right-circularly polarised

(61) LHC=left-circularly polarised

(62) Rx: receiving line

(63) FIG. 8 shows the cross-section of an individual receiving antenna. This is an axial corrugated horn emitter with an integral septum polarizer for a frequency of 76 to 81 GHz. The space requirement of the individual receiving antenna in the horizontal and the vertical direction is 4.2 mm. The aperture opening is specified as 4.1 mm.

(64) FIG. 9 shows the transmitting channel structure of a transmission antenna. The transmitter is fed with a right-circularly and a left-circularly polarised signal according to a temporal sequence.

(65) The symbols indicate as follows here:

(66) RHC=right-circularly polarised

(67) LHC=left-circularly polarised

(68) Timing=temporal sequence of the switching process

(69) FIG. 10 shows the cross-section of a transmission antenna. This is an axial corrugated horn emitter with an integral septum polarizer for a frequency of 76 to 81 GHz. The space requirement of the transmission antenna is 6.3 mm in the horizontal and the vertical direction. The aperture opening is specified as 6.1 mm.

(70) FIG. 11 shows, as an example, the reflection focal points on the vehicle and, with irradiation by an electromagnetic left-circularly polarised wave, their effect upon the polarisation direction of the wave. With a reflection focal point that is in the form of a corner reflector the polarisation direction rotates to the cross-polar, while a reflection focal point that is dihedral in form does not bring about any change to the polarisation direction.

(71) The symbols indicate as follows here:

(72) RHC=right-circularly polarised

(73) LHC=left-circularly polarised

(74) FIG. 12 shows the signal processing of the polarimetric radar. Here 3 signal processing algorithms running in parallel are shown which evaluate the data of the receiving channels for left- and right-circularly polarised signals individually and in combination. The range, speed, angle and object category identification information is obtained by means of the shown arrangement of real and complex Fourier transforms, weighted receiving channel additions and amplitude distribution evaluations according to the principles of half-line transmission location changeover and digital beam shaping.

(75) The symbols indicate as follows here:

(76) RHC: right-circularly polarised

(77) LHC: left-circularly polarised

(78) FFT: Fourier transform

(79) SAR: transmission location changeover

(80) DBF: digital beam shaping/beam steering

(81) FIG. 13 shows the frequency modulation of the transmitted signal over time. The notions of carrier frequency and frequency deviation are implemented in the graph here.

(82) The symbols indicate as follows here:

(83) f: frequency

(84) t: time

(85) f.sub.T: carrier frequency

(86) ?f: frequency deviation