A method of generating a flight path

Abstract

The embodiment of the application provides a track generation method, relates to the technical field of radar systems, and is used for accurately generating the track of a target object in the detection process of a radar system. In the method, first, each transmitting antenna in an antenna array of a radar is controlled to simultaneously transmit a signal, and a phase is added to the signal transmitted by each transmitting antenna; each signal received by each receiving antenna in the antenna array is subjected to fast Fourier transform to obtain an RV graph of each signal; in the RV graph, a plurality of first target points are determined; the plurality of first target points are points with a signal strength greater than a first threshold value; a first track of a first target object is determined according to the plurality of first target points; and the first track is used for representing the trajectory of the first target object.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of radar system, in particular to a track generation method. BACKGROUND

[0002] Millimeter wave radar is gradually used in security perimeter, radar ball fusion and other scenes. These scenes require millimeter wave radar to not only cover a wide range, but also to measure far in all field angles. At present, the method of using digital beam forming or analog beam forming to do electronic scanning can achieve these requirements. In this way, a very wide field angle can be achieved, and far can be measured in these field angles. However, the disadvantage is that the scanning time is long, the frame rate is low, and the generated track is not accurate.

[0003] Therefore, there is a problem of inaccurate track generation in the prior art. SUMMARY

[0004] Embodiments of the present application provide a track generation method, device and electronic equipment to accurately generate the track of a target object in the radar system detection process.

[0005] In a first aspect, the embodiments of the present application provide a track generation method, comprising: controlling each transmitting antenna in an antenna array of a radar to transmit a signal simultaneously, and adding a phase to the signal transmitted by each transmitting antenna; performing fast Fourier transform on each signal received by each receiving antenna in the antenna array to obtain an RV graph of the each signal; determining a plurality of first target points in the RV graph; the plurality of first target points are points with signal strength greater than a first threshold; determining a first track of a first target object according to the plurality of first target points; the first track is used to represent the trajectory of the first target object.

[0006] Based on the above scheme, the radar system can accurately generate the track of the detected target object to achieve accurate measurement of the target object by the radar system.

[0007] In a possible implementation, the determining of the first track of the first target object according to the plurality of first target points comprises: respectively associating the plurality of first target points in two adjacent frames of signals in the time domain to obtain a second track of a plurality of second target points; the plurality of second target points are part or all of the plurality of first target points; associating the plurality of second target points in a single frame of signal to obtain a first target object composed of the plurality of second target points; and merging the second tracks of the plurality of second target points to obtain the first track of the first target object.

[0008] Based on the above scheme, the first target points belonging to the same target point in the two adjacent frames of signals in the time domain are associated, the target points belonging to the same target object in a single frame of signal are associated to determine the target object, and then the multiple target points belonging to the same target object are merged to generate the track of the target object.

[0009] In a possible implementation, the first target points in the two adjacent frames of signals in the time domain are associated respectively to obtain the second tracks of the multiple second target points, including: the first target points satisfying one or more of the following conditions are associated in the two adjacent frames of signals in the time domain to obtain the second tracks of the multiple second target points: in the two adjacent frames of signals in the time domain, the absolute value of the difference of the distances of the first target points is less than or equal to a second threshold value; in the two adjacent frames of signals in the time domain, if the distance of the first target point in the latter frame is greater than the distance of the first target point in the former frame, the ratio of the difference of the distances of the first target points to the frame period of the signal is greater than a third threshold value; if the distance of the first target point in the latter frame is less than or equal to the distance of the first target point in the former frame, the ratio of the difference of the distances of the first target points to the frame period of the signal is less than or equal to a fourth threshold value; in the two adjacent frames of signals in the time domain, the absolute value of the difference of the de-blurring Doppler velocities of the first target points is less than or equal to a fifth threshold value.

[0010] Based on the above scheme, the first target points in the two adjacent frames of signals in the time domain are judged according to the conditions, and the first target points in the two adjacent frames of signals can be accurately associated.

[0011] In a possible implementation, the multiple first target points are associated in a single frame of signal to obtain multiple second target objects composed of the multiple first target points; the first target object is one of the multiple second target objects; if there is no second track of a third target point in the two adjacent frames of signals in the time domain, the third target point is associated with the second target object to which the third target point belongs; and the position information of the second target object to which the third target point belongs in the single frame of signal is used as the starting position information of the second target object to which the third target point belongs.

[0012] In a possible implementation, the first ghost point probability of the multiple first target points is determined according to the azimuth angles and the Doppler velocities of the multiple first target points in the two adjacent frames of signals in the time domain, including: the azimuth angles and the Doppler velocities of the multiple first target points are determined in a single frame of signal; the first ghost point probability of the multiple first target points is determined according to the azimuth angles and the Doppler velocities of the multiple first target points in the two adjacent frames of signals in the time domain; and the first track is updated according to the first ghost point probability.

[0013] Based on the above scheme, the influence of the ghost points caused by the multipath effect in the track can be reduced by calculating the first ghost point probability, thereby improving the accuracy of track generation.

[0014] In a possible implementation, in a single frame signal, a second ghost point probability of the plurality of first target points is determined according to the azimuth angle and the Doppler velocity of the plurality of first target points; in two frame signals adjacent in the time domain, a first ghost point probability of the plurality of first target points is determined according to the second ghost point probability of the plurality of first target points in the single frame signal; wherein the second ghost point probability is determined according to one or more of the following: a radial distance from the radar and a radial distance from the radar to the first target; or a difference between the Doppler velocity and the Doppler velocity of the first target; or an existing time length; or a distance from the first target in the single frame signal; or a heading angle; or a correlation duty cycle; the correlation duty cycle is a ratio of a correlation time to the existing time length, and the correlation time represents a time for judging that the ghost point belongs to a target point of the first target.

[0015] Based on the above scheme, the first ghost point probability can be calculated by judging the second ghost point probability caused by the multipath effect, the influence of the ghost points caused by the multipath effect in the track can be reduced, and the accuracy of track generation can be improved.

[0016] In a possible implementation, in the first track, a first target point of a secondary reflection is determined; and the first track is updated according to the first target point of the secondary reflection.

[0017] Based on the above scheme, the influence of the ghost points caused by the secondary reflection on the accuracy of the track can be removed.

[0018] In a possible implementation, the following operations are performed on any fourth target point in a plurality of fourth target points included in the first track, the plurality of fourth target points being part or all of the plurality of first target points: in a single frame signal, a search space of the any fourth target point is determined according to an azimuth angle standard deviation of the any fourth target point and a distance standard deviation of the any fourth target point in the RV map, and a deambiguating Doppler velocity of the fourth target point is less than or equal to a sixth threshold value; in the single frame signal, an occupied probability of each grid in the search space is determined; in two frame signals adjacent in the time domain, each grid is filtered and fused by using the principle of evidence theory, and for each grid, when the occupied probability of the grid exceeds a seventh threshold value, it is determined that the grid is in an occupied state; and according to the occupied state, an uninterested probability of a target object to which the fourth target point belongs is determined; the uninterested probability represents a probability that the target object to which the fourth target point belongs is an uninterested target.

[0019] Based on the above scheme, the radar system can remove the influence of the non-interesting target in the track by the grid map method to calculate the non-interesting probability, so that the track generation is more accurate.

[0020] In a possible implementation, before the operation of performing the following operation on any fourth target point in the plurality of fourth target points included in the first track, the method further includes: determining, in the single frame signal, that a distance between a position of a target object to which the plurality of fourth target points belong and a starting position of the target object is less than or equal to an eighth threshold value; determining, in the single frame signal, that an absolute speed of the target object is less than or equal to a ninth threshold value; determining that the target object occupies a grid more than or equal to a tenth threshold value in the first frame signal; calculating a position variance of the target object according to positions of the target object in each frame signal; and determining that the position variance of the target object is less than or equal to an eleventh threshold value.

[0021] Based on the above scheme, whether the measured track is a track of a non-interesting target can be determined in combination with the grid map, so that the influence of the non-interesting target on the accuracy of the track is removed, and the accuracy of track generation is improved.

[0022] In a possible implementation, the determining, in the RV map, the plurality of first target points includes: performing peak value detection and ordered constant false alarm rate (OS-CFAR) detection on each point in the RV map in the distance dimension to obtain a first detection result; performing OS-CFAR detection on a point whose first detection result is greater than a twelfth threshold value in the Doppler velocity dimension to obtain a second detection result; and determining a point whose second detection result is greater than or equal to a thirteenth threshold value as a first target point.

[0023] Based on the above scheme, the accuracy of the determined first target point can be ensured by performing OS-CFAR detection in the distance dimension and the Doppler velocity dimension respectively.

[0024] In a possible implementation, the antenna array includes three transmitting antennas and four receiving antennas, the three transmitting antennas are arranged at equal intervals with an interval of 1 Lambda, the lowest part of the middle transmitting antenna and the lowest parts of the two transmitting antennas on the two sides have an interval of 0.5 Lambda, the rightmost transmitting antenna is horizontally spaced apart from the leftmost transmitting antenna by 2 Lambda, and is vertically spaced apart from the leftmost transmitting antenna by 0 Lambda, and the four receiving antennas are arranged at equal intervals with an interval of 0.5 Lambda and a vertical interval of 0 Lambda.

[0025] Based on the above scheme, by controlling the beam direction of the transmitting antenna and reducing the number of transmitted beams, the measurement time is reduced on the basis of realizing large-angle long-distance measurement, and the real-time requirement is met.

[0026] In a second aspect, the embodiments of the present application provide a track generation device, comprising:

[0027] A transceiving unit is configured to transmit and receive signals.

[0028] A processing unit is configured to perform the following operations:

[0029] The processing unit is configured to control each transmitting antenna in an antenna array of a radar to transmit a signal simultaneously, and to attach a phase to the signal transmitted by each transmitting antenna; to perform fast Fourier transform on each signal received by each receiving antenna in the antenna array to obtain an RV map of the signal; to determine a plurality of first target points in the RV map; the plurality of first target points are points with a signal strength greater than a first threshold; and to determine a first track of a first target object according to the plurality of first target points; the first track is used to represent a trajectory of the first target object.

[0030] In a possible implementation, the processing unit is configured to determine a track of a first target object according to the plurality of first target points, and is configured to: associate the plurality of first target points in two adjacent frames of signals in a time domain respectively to obtain a second track of a plurality of second target points; the plurality of second target points are part or all of the plurality of first target points; and associate the plurality of second target points in a single frame of signal to obtain a first target object composed of the plurality of second target points; and combine the second tracks of the plurality of second target points to obtain the first track of the first target object.

[0031] In a possible implementation, the processing unit is configured to determine a track of a first target object according to the plurality of first target points, and is configured to: associate the plurality of first target points in two adjacent frames of signals in a time domain respectively to obtain a second track of a plurality of second target points; the plurality of second target points are part or all of the plurality of first target points; and associate the plurality of second target points in a single frame of signal to obtain a first target object composed of the plurality of second target points; and combine the second tracks of the plurality of second target points to obtain the first track of the first target object. In a possible implementation, the processing unit is configured to determine a track of a first target object according to the plurality of first target points, and is configured to: associate the plurality of first target points in two adjacent frames of signals in a time domain respectively to obtain a second track of a plurality of second target points; the plurality of second target points are part or all of the plurality of first target points; and associate the plurality of second target points in a single frame of signal to obtain a first target object composed of the plurality of second target points; and combine the second tracks of the plurality of second target points to obtain the first track of the first target object.

[0032] In a possible implementation, the processing unit is further configured to: associate the plurality of first target points in a single frame signal to obtain a plurality of second target objects composed of the plurality of first target points; the first target object is one of the plurality of second target objects; if there is no second track of a third target point in two frame signals adjacent in time domain, associate the third target point with a second target object to which the third target point belongs; and take position information of the second target object to which the third target point belongs in the single frame signal as starting position information of the second target object to which the third target point belongs.

[0033] In a possible implementation, the processing unit is further configured to: determine azimuth angles and Doppler velocities of the plurality of first target points in a single frame signal; in two frame signals adjacent in time domain, determine first ghost point probabilities of the plurality of first target points according to the azimuth angles and the Doppler velocities of the plurality of first target points; and update the first track according to the first ghost point probabilities.

[0034] In a possible implementation, the processing unit is further configured to: in a single frame signal, determine second ghost point probabilities according to the azimuth angles and the Doppler velocities of the plurality of first target points; and in two frame signals adjacent in time domain, determine first ghost point probabilities of the plurality of first target points according to the second ghost point probabilities of the plurality of first target points in the single frame signal; wherein the second ghost point probability is determined according to one or more of the following: a radial distance from the radar and a radial distance of the first target object from the radar; or a difference between a Doppler velocity and a Doppler velocity of the first target object; or an existing time length; or a distance from the first target object in a single frame signal; or a heading angle; or an association duty cycle; the association duty cycle is a ratio of an association time to the existing time length, and the association time represents a time for judging that a first ghost point is a target point belonging to the first target object.

[0035] In a possible implementation, the processing unit is further configured to: determine a twice-reflected first target point in the first track; and update the first track according to the twice-reflected first target point.

[0036] In a possible implementation, the processing unit is further configured to: perform the following operations on any fourth target point in a plurality of fourth target points contained in the first track, the plurality of fourth target points being part or all of the plurality of first target points: determine a search space of the any fourth target point according to an azimuth angle standard deviation of the any fourth target point and a distance standard deviation of the any fourth target point in the RV map, the fourth target point having a solution ambiguity Doppler velocity less than or equal to a sixth threshold value; determine an occupancy probability of each grid in the search space in the single frame signal; perform filtering fusion on each grid in two signals adjacent in time domain, and determine that the grid is in an occupied state when the occupancy probability of the grid is greater than a seventh threshold value for each grid according to a principle of evidence theory; and determine an uninterested probability of a target object to which the fourth target point belongs according to the occupied state, the uninterested probability representing a probability that the target object to which the fourth target point belongs is an uninterested target.

[0037] In a possible implementation, before the processing unit performs the following operations on any fourth target point in a plurality of fourth target points contained in the first track, the processing unit is further configured to: determine that a distance between a position of a target object to which the plurality of fourth target points belong and a starting position of the target object is less than or equal to an eighth threshold value in the single frame signal; determine that an absolute velocity of the target object to which the plurality of fourth target points belong is less than or equal to a ninth threshold value in the single frame signal; determine that the target object to which the plurality of fourth target points belong exists in a grid whose occupancy times are greater than or equal to a tenth threshold value in the first frame signal; calculate a position variance of the target object to which the plurality of fourth target points belong according to positions of the target object in each frame signal; and determine that the position variance of the target object to which the plurality of fourth target points belong is less than or equal to an eleventh threshold value.

[0038] In a possible implementation, the processing unit is configured to determine the plurality of first target points in the RV map by: performing peak detection and ordered constant false alarm rate (OS-CFAR) detection on each point in the RV map in a distance dimension to obtain a first detection result; performing OS-CFAR detection on a point whose first detection result is greater than a twelfth threshold value in a Doppler velocity dimension to obtain a second detection result; and determining that a point whose second detection result is greater than or equal to a thirteenth threshold value is a first target point.

[0039] In a possible implementation, the transceiving unit is configured to: the antenna array is composed of three transmitting antennas and four receiving antennas, the three transmitting antennas are arranged at equal intervals with an interval of 1 Lambda, the lowest part of the middle transmitting antenna and the lowest part of the two transmitting antennas on the two sides have an interval of 0.5 Lambda, the leftmost transmitting antenna and the rightmost transmitting antenna have a horizontal interval of 2 Lambda and a vertical interval of 0 Lambda, and the four receiving antennas are arranged at equal intervals with an interval of 0.5 Lambda and a vertical interval of 0 Lambda.

[0040] In a third aspect, an electronic device is provided, including:

[0041] a transceiver configured to receive and send signals;

[0042] a processor connected to the transceiver, configured to process signals received and sent by the transceiver, and implement the method in the first aspect or the second aspect when processing the signals.

[0043] In a fourth aspect, a computer readable storage medium is provided, including:

[0044] The computer readable storage medium stores computer instructions, when the computer instructions run on a computer, the computer instructions make the computer execute the method in any one of the first aspect or the second aspect.

[0045] The above-mentioned aspects of the second aspect to the fourth aspect and the technical effects that can be achieved by the aspects are described above in the first aspect or the various possible schemes in the first aspect, which will not be repeated here. BRIEF DESCRIPTION OF DRAWINGS

[0046] In order to more clearly illustrate the technical solutions of the embodiments of the present application or the prior art, the drawings needed in the embodiment description will be briefly introduced. Obviously, the drawings in the following description are only some embodiments of the present application.

[0047] Figure 1 A communication system schematic diagram to which the track generation method provided by the embodiments of the present application is applicable;

[0048] Figure 2 An exemplary flowchart of a track generation method provided by the embodiments of the present application;

[0049] Figure 3 A directivity diagram of three subframes transmitted by the three transmitting antennas provided by the embodiments of the present application;

[0050] Figure 4An antenna layout diagram provided by the embodiment of the present application;

[0051] Figure 5 A phase shifter calibration flow diagram provided by the embodiment of the present application;

[0052] Figure 6 A transmit waveform diagram provided by the embodiment of the present application;

[0053] Figure 7 An algorithm flow diagram of track generation and update provided by the embodiment of the present application;

[0054] Figure 8 A first ghost point diagram provided by the embodiment of the present application;

[0055] Figure 9 A multipath target point reflection diagram provided by the embodiment of the present application;

[0056] Figure 10 An apparatus diagram provided by the embodiment of the present application;

[0057] Figure 11 A structure diagram of an electronic device provided by the embodiment of the present application. DETAILED DESCRIPTION

[0058] To make the purpose, technical scheme and advantages of the embodiments of the present application clearer, the technical scheme of the present application will be described clearly and completely below with reference to the drawings in the embodiments of the present application. Obviously, the described embodiments are part of the embodiments of the present application, rather than all the embodiments. Based on the embodiments described in the present application, all other embodiments obtained by those skilled in the art without creative labor fall within the scope of protection of the present application.

[0059] The terms "first" and "second" in the embodiments of the present application are used to distinguish different objects, rather than to describe a specific order. In addition, the term "comprises" and any variations thereof are intended to cover non-exclusive protection. For example, a process, method, system, product or device including a series of steps or units is not limited to the listed steps or units, but optionally includes steps or units not listed or optionally includes other steps or units inherent to the process, method, product or device. "Multiple" in the present application can mean at least two, for example, two, three or more, and the embodiments of the present application are not limited.

[0060] In addition, the term "and / or" in this document is only used to describe the association relationship of the associated objects, which means that there can be three relationships, for example, A and / or B can represent three cases of A alone, A and B together, and B alone. In addition, the character " / " in this document generally represents an "or" relationship between the front and rear associated objects without special description.

[0061] In the prior art, in order to achieve large-angle and long-distance measurement, a radar needs a long measurement time and a low frame rate, which is difficult to meet the real-time requirement, and there is usually interference of static targets such as grass and trees during measurement, which affects the accuracy of measurement. Therefore, how to reduce the measurement time of the radar while achieving large-angle and long-distance measurement and improve the accuracy of measurement is an urgent problem to be solved.

[0062] Therefore, the embodiment of the present application provides a track generation method, in which the direction of the beam emitted by the transmitting antenna is controlled and the number of beams emitted is reduced, so as to reduce the measurement time on the basis of achieving large-angle and long-distance measurement and meet the real-time requirement. The influence of uninterested targets such as grass and trees is removed by the grid map learning method, so as to improve the accuracy of radar measurement.

[0063] Referring to Figure 1 The track generation method provided by the embodiment of the present application is applicable to a radar system 100. The radar system 100 includes a processor 101 and an antenna 102. The antenna 102 is used for transmitting signals and receiving reflected signals. The processor 101 is used for controlling the transmitting antenna to transmit signals and processing the signals received by the receiving antenna, so as to realize detection of the position of a target object and generation of the track of the target object.

[0064] Referring to Figure 2 An exemplary flowchart of a track generation method provided by the embodiment of the present application is shown in FIG. 2. The method can be applied to a radar system and includes the following operations:

[0065] S201: The processor of the radar system controls each transmitting antenna in the antenna array of the radar to transmit signals simultaneously, and adds a phase to the signal transmitted by each transmitting antenna.

[0066] Before the processor of the radar system controls each transmitting antenna in the antenna array of the radar to transmit signals simultaneously, the transmitted signals are subjected to waveform modulation. Each frame of signal contains three sub-frames, and each sub-frame can include 128 Chirp signals.

[0067] Specifically, the processor can control the three transmitting antennas to respectively transmit the three sub-frames. For example, the processor can control the first transmitting antenna to transmit the first sub-frame of the first frame signal, control the second transmitting antenna to transmit the second sub-frame of the first frame signal, and control the third transmitting antenna to transmit the third sub-frame of the first frame signal. It should be noted that the angles pointed by the sub-frames transmitted by the three transmitting antennas are different, so as to realize larger angle radiation by fewer beams.

[0068] Referring to Figure 3 The three sub-frames transmitted by the three transmitting antennas provided in the embodiment of the present application are provided with direction patterns. In order to realize that the angles pointed by the sub-frames transmitted by the three transmitting antennas are different, the signals transmitted by each transmitting antenna need to be attached with a phase. The embodiment of the present application takes the first sub-frame of a frame signal pointing to 0 degree, the second sub-frame of a frame signal pointing to 19 degrees, and the third sub-frame of a frame signal pointing to -19 degrees as an example for description. That is, the additional phase of the first transmitting antenna can be set to 0, the additional phase of the second transmitting antenna can be set to 2*pi*sin(Angle), and the additional phase of the third transmitting antenna can be set to 2*pi*2*sin(Angle), where Angle is 19 degrees.

[0069] In a possible implementation, the radar system sets the value of the phase shifter of each transmitting antenna according to the phase shifter calibration parameter table through the principle of digital beam forming, attaches a phase to each transmitting antenna, and thus controls the sub-frames of the signals transmitted by each transmitting antenna to point to different angles.

[0070] S202: The processor of the radar system performs fast Fourier transform on each signal received by each receiving antenna in the antenna array, and obtains an RV diagram of each signal.

[0071] The signal received by the receiving antenna is the signal transmitted by the transmitting antenna in the foregoing S201. The radar system detects the position and motion track of an object by receiving the signal reflected by the object.

[0072] After the signal is received by the receiving antenna, the processor first samples each Chirp received by each receiving antenna through an Analog to Digital Converter (ADC), and then performs distance dimension spectrum analysis on each Chirp of each receiving antenna through a distance dimension Fast Fourier Transform (FFT) after frequency compensation is performed on each Chirp of each receiving antenna. Then, the result of the distance dimension FFT is stored in a complex number array array[M][N][K]. Wherein M is the number of sampling points of each Chirp, N is the number of Chirps of each subframe of each receiving antenna, and K is the index number of the receiving antenna. The values of M, N and K can be set according to actual conditions, and the present application does not limit this. In the embodiment of the present application, the number of Chirps of each subframe of each receiving antenna is set to 128, so M is 128, the number of sampling points of each Chirp is set to 128, so N is 128, and since the radar system provided in the embodiment of the present application includes four receiving antennas, K = 0, 1, 2, 3.

[0073] The processor performs a Doppler velocity dimension FFT on array[M][N][K] according to each receiving antenna, respectively, to obtain an R-V diagram of each signal.

[0074] S203: The radar system processor determines a plurality of first target points in the RV diagram.

[0075] According to the RV diagram obtained in S202, a peak value detection and an Ordered Statistics-Constant False Alarm Rate (OS-CFAR) detection are performed in the distance dimension to obtain a first detection result, and an OS-CFAR detection is continuously performed on the points greater than a twelfth threshold value in the Doppler velocity dimension, and the points greater than a thirteenth threshold value are considered as the first target points. Wherein the distance value of the first target point is the product of the distance dimension index and the distance resolution, the Doppler velocity value of the target point is the product of the Doppler velocity dimension index and the Doppler velocity resolution, the product of the distance dimension index and the distance resolution, the Doppler velocity dimension index and the Doppler velocity resolution are all preset according to actual conditions, the twelfth threshold value and the thirteenth threshold value are both set according to the distance unit and actual conditions, the twelfth threshold value can be the same as or different from the thirteenth threshold value, and the present application does not limit this.

[0076] S204: The radar system processor determines a first track of a first target object according to the plurality of first target points.

[0077] The first target object is determined and the track of the first target object is generated in combination with the first target points determined in S203 and the position information and Doppler velocity of the first target points. Wherein the target object is determined by the first target points.

[0078] Specifically, the antenna array of the radar system is composed of three transmitting antennas and four receiving antennas. Referring to Figure 4 The antenna array layout provided by the embodiment of the present application is shown in the figure, and the transmitting antennas are marked as Tx and the receiving antennas are marked as Ry. The three transmitting antennas are T0, T1 and T2 respectively. As shown in the figure, the horizontal distance between T0 and T1 is 1Lambda, and the vertical distance is 0.5Lambda. The horizontal distance between T0 and T2 is 2Lambda, and the vertical distance is 0Lambda. The four receiving antennas are R0, R1, R2 and R3 respectively, and the horizontal distance between any two adjacent receiving antennas is 0.5Lambda, and the vertical distance is 0Lambda. The distance between the three transmitting antennas and the four receiving antennas, that is, the distance between R3 and T0 in the middle, can be set according to the size of the radio frequency board. For example, a larger distance is set when the radio frequency board is larger, and a smaller distance is set when the radio frequency board is smaller, which is not limited in the present application. Figure 1

[0079] It should be noted that before the transmitting antenna transmits the signal, the processor of the radar system calibrates the phase shifter of the transmitting antenna. The phase shifter calibration parameter table is obtained by calibrating the phase shifter of the radar system, as shown in Figure 5 The phase shifter calibration process provided by the embodiment of the present application is shown in the figure, which can include the following processes:

[0080] S501, control the environmental temperature to be-40 degrees, and let i=1.

[0081] Fix the radar whole machine with radome at a certain position, and then control the environmental temperature to be-40 degrees. Set i=1, where i represents the i-th transmitting antenna.

[0082] S502, set the angle of reflection.

[0083] Set an angle of reflection with a radar reflection cross-sectional area of 1dbsm as a target, place the target at a position 5m away from the radar whole machine, and use a laser to align so that the angle of the target relative to the radar normal is 0 degrees.

[0084] S503, set the value of the i-th transmitting antenna phase shifter to 0.

[0085] S504, the transmitting antenna continuously transmits 64 Chirps.

[0086] Under the value of the phase shifter, control the transmitting antenna to continuously transmit 64 Chirps.

[0087] S505, record the sampling points of the first receiving antenna for 64 Chirps.

[0088] ​S506, the phase of the angle echo is calculated.

[0089] The received data of the first receiving antenna is subjected to 2D Fourier transform, and the phase of the angle echo is calculated. The phase of the angle echo is obtained by searching the position of the angle echo in the two-dimensional spectrum according to the position information and the Doppler velocity of the angle echo.

[0090] S507, the calculated phase is subtracted by the phase of the first transmitting antenna when the phase shifter is 0.

[0091] The phase calibration value of the i-th transmitting antenna under the current temperature and the current phase shifter value is obtained.

[0092] S508, it is judged whether the temperature is equal to 120 degrees.

[0093] If the temperature is not equal to 120 degrees, S509 is executed. If the temperature is equal to 120 degrees, S510 is executed.

[0094] S509, the temperature is increased by 20 degrees.

[0095] After the ambient temperature is increased by 20 degrees, S504-S508 are repeatedly executed.

[0096] S510, it is judged whether the phase shifter value is equal to the maximum value.

[0097] Specifically, the embodiment of the present application takes a 6-bit phase shifter as an example, and the phase shifter value needs to be set 64 times, i.e. the value is from 0 to 63. That is, it is judged whether the phase shifter value is greater than 63. If the phase shifter value is equal to 63, S512 is executed. If the phase shifter value is not equal to 63, S511 is executed.

[0098] S511, the phase shifter value is increased by 1.

[0099] Then, S504-S510 are repeatedly executed.

[0100] S512, it is judged whether i is equal to 3.

[0101] If i is equal to 3, S514 is executed. If i is not equal to 3, S513 is executed.

[0102] S513, i=i+1.

[0103] Then, S503-S512 are repeatedly executed.

[0104] S514, the phase shifter calibration parameter table is obtained.

[0105] Through the above process, the phase shifter calibration parameter table can be obtained, which is used to find the phase calibration value of each transmitting antenna at different temperatures.

[0106] In a possible implementation, the processor can also perform waveform modulation on each frame in the transmitted signal in a random delay manner to achieve the additional phase for each transmitting antenna. See Figure 6 A transmitted waveform diagram provided by an embodiment of the present application. Figure 5 In the formula, T is the period of each Chirp. δ is a random delay equal to k*Δt, where k is a random number and Δt is a sampling interval. δ and δ2 can be the same value or different values.

[0107] Optionally, to solve the influence of the random delay method used in waveform modulation and enhance the anti-interference capability of the radar system, phase compensation can be performed on each Chirp after the spectrum analysis in the range dimension, and the compensation phase is exp(-j*2*pi*RangeIdx*k / N). In the formula, RangeIdx is the range dimension index, k is the random number of each Chirp, and N is the ADC sampling number of each Chirp. The value of the range unit represents the distance of each sampling point to the coordinate origin, and the number of range units is equal to the number of sampling points.

[0108] When the radar system has good range resolution and Doppler velocity resolution, it can be considered that there is only one first target point under the same distance and the same speed. Therefore, according to the range unit identifier and the Doppler velocity unit identifier of each receiving antenna, the first target point with the same distance and the same speed in the above target points is found, and the Digital Beam Forming (DBF) processing in the azimuth dimension is performed on these first target points to obtain a power spectrum, the maximum value in the power spectrum is found, and the azimuth angle of the first target point can be calculated through the relationship between the power and the azimuth angle and the compensation value of the amplitude and phase of the Tx-Ry pre-stored, where x is the transmitting antenna number and y is the receiving antenna number.

[0109] In the formula, the calculation method of the compensation value of the amplitude and phase of Tx-Ry is as follows: an angle reflector with a reflection cross-sectional area of 1dbsm is placed at a position 5m away from the radar system, then the angle reflector is experimented every 1 degree in the range of-60 degrees to 60 degrees, the amplitude and phase at the position corresponding to the angle reflector are calculated through the FFT spectrum analysis on the signals obtained by each group of Tx-Ry in the time division multiple input multiple output state in each experiment, and then compared with the pre-stored theoretical value at the angle, and the compensation value of the amplitude and phase of each group of Tx-Ry at the angle is stored.

[0110] In a possible implementation, to improve the measurement accuracy in the case of good signal-to-noise ratio, the azimuth angle of the calculated target point can be interpolated through a quadratic peak interpolation algorithm or a parabolic interpolation algorithm.

[0111] Optionally, the interference of static objects such as grass or trees can be removed by determining the stationary target points in the first target points. Specifically, when K = 0, 1, 2, 3, the array [M] [N] [K] is subjected to 128-point mean filtering along the Doppler direction, and the filtered result is saved to the array array_Average[M][K]. Then the obtained array_Average[M][K] is subjected to ordered constant false alarm (OS-CFAR) processing for each receiving antenna, and the points greater than the discrimination threshold of the OS-CFAR are taken as the stationary target points. The discrimination threshold of the OS-CFAR is preset according to the actual situation and experience, which is not limited in the present application.

[0112] Optionally, the multi-target judgment under the same distance and the same speed can also be performed by a super-resolution algorithm such as a generalized likelihood ratio test (GLRT) criterion and a deterministic maximum likelihood (DML) to calculate the azimuth angles of the multiple first target points under the same distance and the same speed.

[0113] In another possible implementation, the data sampled by the ADC can be subjected to interference removal processing. Specifically, gradient detection can be used to identify some points of sudden increase or decrease in amplitude on the time-domain waveform, and then a band-pass filter can be used for filtering processing. For example, the data subjected to the band-pass filter filtering processing only remains the frequency components in a certain frequency range, and the frequency components in other ranges are completely attenuated.

[0114] In summary, the processor can obtain the distance and the azimuth angle of the first target points. The distance represents the distance of the target point relative to the radar system, and the azimuth angle represents the angle of the target point relative to the normal line of the radar.

[0115] Optionally, since there can be noise in the signal received by the receiving antenna, there can be noise points in the above target points, and therefore the radar system can determine the effective target points from the first target points. The signal-to-noise ratio (SNR) of the noise points is less than the SNR threshold. The SNR threshold is preset according to experience and actual situation, which is not limited in the present application.

[0116] Firstly, SNR of each first target point is calculated. The SNR is the ratio of the signal power of the target point to the noise power. The signal power of the first target point can be calculated by the value in the power spectrum. The noise power can be obtained by the following method: firstly, each first target point in the R-V diagram is sorted in ascending order according to the power, and the Kth value is selected as the short-time noise value. The long-time noise value of each receiving antenna can be obtained by low-pass filtering the short-time noise value. Finally, the maximum value of the noise in the long-time noise values calculated by all receiving antennas is selected as the noise power. Wherein, K is set according to experience and actual situation, and the present application does not limit it. When the signal-to-noise ratio of the target point is greater than the SNR threshold, the first target point is determined as an effective target point.

[0117] A possible implementation, the transmission time of each subframe can be 8ms, and the time for receiving signals and processing is 12ms. Thus, three beams are performed for 60ms, and the information of the target points detected by the radar can be output, and the real-time requirement of the radar system can be realized.

[0118] Referring to Figure 7 The algorithm flowchart provided by the embodiment of the present application for generating and updating the track can include the following flowchart:

[0119] S701, in the time domain, the processor respectively associates a plurality of first target points in two adjacent frames of signals to obtain a second track of a plurality of second target points.

[0120] In order to generate the track of the first target object, firstly, the first target points determined in the time domain of the two adjacent frames of signals are respectively associated, and the first target points belonging to the same target point in the first target points of the two adjacent frames of signals are determined as the second target points. When the first target points in the current and next two frames of signals satisfy the following conditions, the first target points in the current and next two frames of signals are associated to obtain the second track of the plurality of second target points:

[0121] Condition 1, in the time domain, the absolute value of the difference of the distances of the first target points in the two adjacent frames of signals is less than or equal to the second threshold. Specifically, the absolute value of the difference between the distance of the first target point of the next frame and the distance of the first target point of the previous frame satisfies the formula |RC2-RC|≤gate. Wherein, gate=Vmax*T, RC2 is the distance of the first target point of the next frame, RC is the distance of the first target point of the previous frame, Vmax is the preset maximum allowable speed of the first target point, T is the frame period, and gate is the second threshold.

[0122] Condition 2, if the distance of the first target point in the next frame is greater than the distance of the first target point in the previous frame, the ratio of the difference value of the distance of the first target point to the frame period of the signal is greater than a third threshold value; if the distance of the first target point in the next frame is less than or equal to the distance of the first target point in the previous frame, the ratio of the difference value of the distance of the first target point to the frame period of the signal is less than or equal to a fourth threshold value. Specifically, when RC2>RC, it satisfies (RC2-RC) / T>(VC+VC2) / 2; when RC2≤RC, it satisfies (RC2-RC) / T≤(VC+VC2) / 2. Wherein, VC is the Doppler velocity of the first target point in the last frame, VC2 is the Doppler velocity of the first target point in the current frame, the third threshold value and the fourth threshold value are both (VC+VC2) / 2.

[0123] Condition 3, in the two adjacent frames of signals in the time domain, the absolute value of the difference of the deblurring Doppler velocity of the first target point is less than or equal to a fifth threshold value. Specifically, the deblurring Doppler velocity of the first target point in the previous frame and the deblurring Doppler velocity of the first target point in the next frame satisfy |VO2-VO|≤Amax*T. Wherein, VO2 is the deblurring Doppler velocity of the first target point in the current frame, VO is the deblurring Doppler velocity of the first target point in the last frame, Amax is the maximum allowable acceleration of the first target point, Amax*T is the fifth threshold value, and satisfies Amax*T=|Ve2-Ve1| / 2, Ve1 is the Doppler velocity unblurring period of the last frame, and Ve2 is the Doppler velocity unblurring period of the current frame.

[0124] Specifically, the deblurring Doppler velocity of the first target point is calculated by the following method: in the case of Ve1<Ve2, the blurring coefficient k in a single frame signal is calculated as k=mod(VC-VC2,Ve1) / (Ve2-Ve1), and the time interval when k is in the interval (-1, 1) is the unblurring interval, and at this time the Doppler velocity of the first target point is the deblurring Doppler velocity. When k>1, the compensation amount of the Doppler velocity of the first target point is calculated by calculating the Doppler blurring coefficient K=k-ve1 / (ve2-ve1), so as to obtain VO and VO2.

[0125] It should be noted that before the correlation of the plurality of first target points in the two adjacent frames of signals in the time domain, the position information of the first target point in the polar coordinates needs to be converted into the position information in the Cartesian coordinates, and the position variance of the first target point in the Cartesian coordinates is calculated based on the linearization principle of the nonlinear function.

[0126] After the correlation of the first target points satisfying the above conditions, the processor can generate the track of the first target point. If all the first target points do not satisfy the above conditions, the processor determines that the track does not exist.

[0127] In a possible implementation, when the first target point in the two adjacent frames of signals satisfying the above condition is associated, the first target point satisfying the above condition can be optimally matched to reduce the amount of calculation. The optimal matching process is as follows: first, the azimuth angle is considered as a certain distribution for weight calculation, and the global nearest neighbor algorithm (GNN) is used for optimal matching. Then, the association hypothesis is made through the multi-target tracking hypothesis algorithm, and the optimal hypothesis is determined by calculating the hypothesis probability of the subsequent several frames of signals.

[0128] In S702, the processor determines whether the track exists.

[0129] If the processor determines that the track does not exist, S703 is performed. If the processor determines that the track exists, S704 is performed.

[0130] In S703, the processor initializes the starting position of the track.

[0131] When the processor determines that the track does not exist, the second track of the third target point is determined, the third target point belonging to the second target object is determined through the third target point in the single frame of signals, and the starting position of the track of the second target object is found according to the related information of the second target object. The third target point is the first target point in the two adjacent frames of signals in the time domain that does not satisfy the association condition.

[0132] Specifically, the second target object is determined by using a representative density-based clustering method (DBSCAN) to perform density association on the first target point in the single frame of signals. First, the association radius and the association density of the DBSCAN algorithm need to be set, wherein the association density is set to 1 in the embodiment of the application. It should be understood that the association density is preset according to the actual situation and experience, which is not limited in the application.

[0133] Since the association of the first target point in the single frame of signals needs to satisfy the following conditions: the square of the longitudinal position difference between the two first target points in the single frame of signals is less than or equal to the sum of the longitudinal variances of the two first target points, and is less than or equal to the longitudinal position fluctuation threshold; the square of the lateral position difference between the two first target points in the single frame of signals is less than or equal to the sum of the lateral variances of the two first target points, and is less than or equal to the lateral position fluctuation threshold. At the same time, the speed difference between the two first target points is less than or equal to the speed fluctuation threshold. The longitudinal position fluctuation threshold and the lateral position fluctuation threshold are both preset according to the distance resolution of the radar system and the angle resolution of the radar system, and the speed fluctuation threshold is preset according to the actual situation and experience, which is not limited in the application.

[0134] Therefore, the correlation radius can be set according to the longitudinal position fluctuation threshold, the lateral position fluctuation threshold and the speed fluctuation threshold. Alternatively, the correlation radius can be set as the sum of the longitudinal position fluctuation threshold, the lateral position fluctuation threshold and the speed fluctuation threshold, which is not limited in the present application.

[0135] The determined correlation density, correlation radius and first target point are input into the DBSCAN clustering algorithm, and the second target object can be determined by clustering.

[0136] Then, the position, speed, radar cross section (RCS) and corresponding variance of the second target object are determined. The specific algorithm is as follows:

[0137] First, the longitudinal velocity difference VDiff1, the lateral velocity difference VDiff2, the oblique positive 45-degree direction velocity difference VDiff3 and the oblique negative 45-degree direction velocity difference VDiff4 of the third target point are calculated. The longitudinal direction represents that the heading angle of the third target point is 0 degrees or 180 degrees, the lateral direction represents that the heading angle of the third target point is 90 degrees or -90 degrees, the oblique positive 45-degree direction represents that the heading angle of the third target point is 45 degrees or -135 degrees, and the oblique negative 45-degree direction represents that the heading angle is -45 degrees or 135 degrees.

[0138] Specifically, VDiff1 = |VO / cos(ф1) - VO2 / cos(ф2)|, if the value of cos(ф1) or cos(ф2) is zero, VDiff1 takes a relatively large value; VDiff2 = |VO / sin(ф1) - VO2 / sin(ф2)|, if the value of sin(ф1) or sin(ф2) is zero, VDiff2 takes a relatively large value; VDiff3 = |VO / cos(ф1-45°) - VO2 / cos(ф2-45°)|, if the value of cos(ф1-45°) or cos(ф2-45°) is zero, VDiff3 takes a relatively large value; VDiff4 = |VO / cos(ф1+45°) - VO2 / cos(ф2+45°)|, if the value of cos(ф1+45°) or cos(ф2+45°) is zero, VDiff4 takes a relatively large value. Wherein, ф1 is the angle of the first target point relative to the coordinate system in the previous frame, and ф2 is the angle of the first target point relative to the coordinate system in the current frame.

[0139] Then the minimum value of VDiff1, VDiff2, VDiff3, VDiff4 is selected, and it is determined whether the minimum value is greater than a set threshold. If VDiff1 is the minimum value and greater than the set threshold, it is considered that the heading angle of the third target point is longitudinal, and it is determined whether the heading angle is 0 degree or 180 degree by using the Doppler sign, and then the speed is calculated. If VDiff2 is the minimum value and greater than the set threshold, it is considered that the heading angle of the third target point is transverse, and it is determined whether the heading angle is 90 degree or negative 90 degree by using the Doppler sign, and then the speed is calculated. If VDiff3 is the minimum value and greater than the set threshold, it is considered that the heading angle of the third target point is diagonal to the positive 45 degree direction, and it is determined whether the heading angle is 45 degree or negative 135 degree by using the Doppler sign, and then the speed is calculated. If VDiff4 is the minimum value and greater than the set threshold, it is considered that the heading angle of the third target point is diagonal to the negative 45 degree direction, and it is determined whether the heading angle is negative 45 degree or 135 degree by using the Doppler sign, and then the speed is calculated. If VDiff1, VDiff2, VDiff3, VDiff4 are all less than the set threshold, the position information of the third target point in the adjacent two frames of signals is comprehensively analyzed, and the heading angle and the speed value are determined.

[0140] Finally, all the third target points are clustered with weights, and the position, speed, RCS and corresponding variance of the second target object and other related information can be obtained. Then, the starting position of the second target object track can be initialized by using the related information of the second target object, and the corresponding Kalman filter parameters and RCS information can be obtained. The clustering method can be selected according to specific conditions, which is not limited in the present application.

[0141] Specifically, the RCS is calculated by the radar equation, which satisfies the following formula: 10log(RCS) = 40log(R) + 10log(Pr / (G*Pt)) + 30log(4*pi) + 10log(Lt) + 10log(Lr) - 10log(Gt*Gr) - 20log(Lambda), wherein R is the distance of the target point or target object from the radar, G is the power amplification parameter, Pr is the power after FFT, Pt is the transmit power, Lt is the energy attenuation coefficient of the transmitting antenna, Lr is the energy attenuation coefficient of the receiving antenna, Gt is the transmitting antenna gain, Gr is the receiving antenna gain, and Lambda is the wavelength of the transmitted electromagnetic wave in free space.

[0142] The variance is calculated by the following method. Firstly, the calculation of the variance can be divided into the variance of the distance measurement, the variance of the Doppler velocity measurement and the variance of the azimuth angle measurement. Since the radar detection is currently based on the signal-to-noise ratio detection criterion, the variance of the radar system detection result is related to the signal-to-noise ratio of the measurement value. The relationship between the variance of the radar system detection result and the signal-to-noise ratio of the measurement value can be represented by a look-up table (LUT) established in advance under laboratory conditions.

[0143] Then the variance of the corresponding measurement value is found by measuring the signal-to-noise ratio of the measurement value. For example, the variance of the distance measurement can be represented as Var(R) = RangeResolution*LUT_Range(SNR), the variance of the Doppler velocity measurement can be represented as Var(V) = VResolution*LUT_Doppler(SNR), wherein RangeResolution represents the distance resolution, VResolution represents the Doppler velocity resolution, LUT_Range(SNR) is a value found in the LUT according to the SNR of the distance measurement value, and LUT_Doppler(SNR) is a value found in the LUT according to the SNR of the Doppler velocity measurement value.

[0144] The variance of the azimuth angle measurement can be represented as Var(Sin(Angle)) = Resolution*LUT_Azimuth(SNR), wherein Angle is the azimuth angle, Sin(Angle) is the sine value of the azimuth angle, Resolution is the azimuth angle resolution, and LUT_Azimuth(SNR) is a value found in the LUT according to the SNR of the azimuth angle measurement value. Since an interpolation algorithm is used in the calculation of the azimuth angle, the variance of the interpolated Sin(Angle) is calculated in combination with the specific interpolation algorithm used, and then the variance of the azimuth angle measurement is calculated by using the linearization principle of the nonlinear function.

[0145] It should be noted that the parameters used in the calculation of the RCS and the variance are pre-stored, including: 1) the measured gain pattern after digital beamforming of three transmitting antennas, which includes the measured gain pattern after three times of digital beamforming with a pointing direction of zero degrees, 19 degrees and -19 degrees in the embodiment of the present application; 2) the measured antenna gain pattern of the receiving antenna; 3) the energy attenuation coefficient under different weather conditions; and 4) the power amplification parameter. When each sub-frame is measured, the power received by the receiving antenna will pass through the power amplifier, so that the FFT gain and other parameters are amplified. The multiple of the FFT gain and other parameters can be regarded as a constant, which is referred to as a power amplification parameter in the embodiment of the present application.

[0146] S704, the processor generates a first track of the first target object.

[0147] When the processor judges that the track exists, firstly, the plurality of second target points are associated in the single frame signal to obtain a first target object composed of the plurality of second target points.

[0148] The first target object is determined by the processor through JPDA or PDA clustering algorithm on the second target points associated with the second track. Optionally, the rear center point of the target object can be taken as the tracking point of the target object track, and the position information and speed and other related information of the tracking point can be taken as the position information and speed and other related information of the target object. Since the same target object can have multiple target points, that is, the same target object can have multiple tracks. In this case, the existence time of the second track of all second target points constituting the first target object and the RCS of all second tracks can be merged through Bayesian fusion to generate the first track of the first target object. It should be understood that the position of the tracking point can be selected according to experience or actual situation, which is not limited in the present application.

[0149] Then, the position coordinates, speed, acceleration and variance and other information of the first track are determined by using Kalman filtering principle. Firstly, the measurement variance matrix R in Kalman filtering is set according to the variance of the first target object, and the system noise matrix Q in Kalman filtering is set according to the difference between the predicted value and the measured value of the related information of the first target object in the plurality of frame signals. Then, the Kalman filtering is performed according to the set R and Q, and the position, speed, acceleration and variance and other related information of the first track can be obtained.

[0150] Specifically, the second target point associated with the track is determined by the following method:

[0151] Firstly, the coordinate system of the first target point is converted, that is, the first target point is transformed into a coordinate system with the heading angle of the track as the X axis (longitudinal direction) and the Y axis (lateral direction) at a 90-degree angle with the X axis, and the coordinates and variance of the first target point are converted.

[0152] Then, the preliminary judgment of association is made based on the position relationship between the first target point and the second track. Let the coordinate of the first target point after conversion of the coordinate system be (X, Y), and the position of the second track be (XO, YO). If |X-XO|<MaxLength and |Y-YO|<MaxWidth, then the preliminary judgment is association, and the next step of accurate position association is continued. Wherein, MaxLength is a maximum length threshold, and MaxWidth is a maximum width threshold. MaxLength and MaxWidth are both preset according to the distance resolution and angle resolution of the radar system, which are not limited in the present application.

[0153] The method for judging the accurate position association is: if X is greater than XO, judging whether the square of (X-XO) is less than or equal to the forward threshold value. If the square of (X-XO) is greater than the forward threshold value, the first target point is not associated with the second track. If the square of (X-XO) is less than or equal to the forward threshold value, judging whether the square of (Y-YO) is less than or equal to the lateral threshold value.

[0154] If X is less than or equal to XO, judging whether the square of (X-XO) is less than or equal to the backward threshold value. If the square of (X-XO) is greater than the backward threshold value, the first target point is not associated. If the square of (X-XO) is less than or equal to the backward threshold value, judging whether the square of (Y-YO) is less than or equal to the lateral threshold value. The forward threshold value, the backward threshold value and the lateral threshold value are all preset according to the variance of the first target point after transformation, the track variance and the position fluctuation threshold value, which are not limited in the present application.

[0155] If the square of (Y-YO) is less than or equal to the lateral threshold value, judging that the first target point is associated with the second track in position, and then judging whether the difference between the Doppler velocity of the second track and the Doppler velocity of the first target point is less than the Doppler velocity threshold value. The Doppler velocity threshold value is preset according to the preset maximum allowable acceleration and the frame period of the signal, which is not limited in the present application. If the square of (Y-YO) is greater than the lateral threshold value, judging that the first target point is not associated with the second track.

[0156] If the difference between the Doppler velocity of the second track and the Doppler velocity of the first target point is less than the Doppler velocity threshold value, judging that the first target point is associated with the second track.

[0157] The above judging steps are repeatedly executed until all the first target points are judged.

[0158] S705, the processor learns the grid map.

[0159] The processor can judge the probability of the target point being an uninterested target such as grass and trees through the method of learning the grid map. The grid map refers to dividing the detection area of the radar system into multiple grids. For example, the area detected by the radar is a 5m*5m area, which can be divided into multiple 10cm*10cm grids. In the embodiment of the present application, the grid map is a polar coordinate grid map, that is, the longitudinal coordinate of the grid map coordinate system is the distance and the lateral coordinate is the azimuth.

[0160] Firstly, the search space of each fourth target point of the polar coordinate grid map is determined. The search space of the polar coordinate grid map is determined according to the distance standard deviation σ r and the azimuth standard deviation σ a of the fourth target point. Specifically, the starting value of the distance of the search space of each fourth target point is r start = range-3* σ r, and if r start is less than zero, r start takes the value of zero; the ending value of the distance is r end = range+3* σ r, and if r end exceeds the maximum distance, r end is the maximum distance, and range is the distance of the fourth target point; the starting value of the azimuth is a start = azimuth-3* σ a, and if a start is less than the minimum azimuth, a start is the minimum azimuth; and the ending value of the azimuth is a end = azimuth+3* σ a, and if a end exceeds the maximum azimuth, a end is the maximum azimuth, and azimuth is the azimuth of the target point. The maximum distance, the minimum azimuth and the maximum azimuth are all preset according to actual conditions, and the present application does not limit them.

[0161] Wherein, the fourth target point is the target point in the first target point whose Doppler velocity after deblurring is less than the first threshold value. Wherein, the threshold value depends on the maximum Doppler velocity that can be measured by the trees or grasses in the actual process.

[0162] After the search space of each fourth target point is determined, each grid in the range is processed as follows: the occupancy of each frame of signal grid is detected, the occupancy times of each grid are calculated, if the grid occupancy times are zero, the SNR of the grid is initialized as the SNR of the fourth target point, the occupancy times are 1, and the continuous non-occupancy times are zero. If the grid occupancy times are greater than zero, the average SNR of the grid in all frames of signals is calculated, the occupancy times are added by 1, and the continuous non-occupancy times are zero. It should be understood that the grid occupancy times are the number of times that the fourth target point exists in the grid in multiple frames of signals, and the continuous non-occupancy times are the number of times that the fourth target point does not exist in the grid in continuous frames of signals.

[0163] For example, in the first frame of signal, the fourth target point does not exist in the grid, so the continuous non-occupancy times of the grid are 1. In the second frame of signal, if the fourth target point does not exist in the grid, the continuous non-occupancy times of the grid are added by 1; if the fourth target point exists in the grid, the continuous non-occupancy times of the grid are 0. The judgment in the second frame of signal is continued for the next frame of signal, and when the judgment for all frames of signals is completed, the continuous non-occupancy times of the grid are obtained.

[0164] After the above processing is performed on each fourth target point, the following processing is performed on the grid in the grid map that has not been processed: if the number of consecutive non-occupations of the grid is less than a threshold value, the number of consecutive non-occupations is incremented by 1; otherwise, the SNR of the grid is set to 0, and the occupation number is set to 0. The threshold value is preset according to experience or actual conditions, and is not limited in the present application.

[0165] S706, the processor updates the first track.

[0166] Updating the first track includes: after the processor performs the grid map learning of S705, removing the non-interested targets and the ghost points by calculating the probability of the track of the non-interested targets such as grass and trees, and the probability of the first ghost point and the second ghost point generated by the processor according to the multipath effect.

[0167] The method for calculating the probability that the first track is not the track of the non-interested targets such as grass and trees is as follows: first, the distance between the position of the target to which the fourth target point belongs in a single frame signal and the starting position of the target to which the fourth target point belongs is determined, and if the distance is greater than an eighth threshold value, the first track is not the track of the non-interested targets, and the confidence that the target to which the fourth target point belongs is not a non-interested target is set to 1. If the distance is less than or equal to the eighth threshold value, it is further determined whether the absolute speed of the target to which the fourth target point belongs is greater than a ninth threshold value, and if the absolute speed of the target to which the fourth target point belongs is greater than the speed threshold value, the target to which the fourth target point belongs is not a non-interested target, and the confidence that the target to which the fourth target point belongs is not a non-interested target is set to 1. If the absolute speed of the target to which the fourth target point belongs is less than or equal to the speed threshold value, the position variance of the target to which the fourth target point belongs is calculated in combination with the position of the target to which the fourth target point belongs in each frame signal.

[0168] In the two adjacent frames of signals in time domain, the principle of evidence theory is used to filter and fuse each grid. For each grid, when the occupancy probability of the grid exceeds the seventh threshold value, it is determined that the grid is in an occupied state. Specifically, whether the starting position of the target object to which the fourth target point belongs exists a grid in an occupied state is judged by the polar coordinate grid map. If not, the confidence that the target object to which the fourth target point belongs is not an uninterested target is set to 0.7. If so, whether the position variance of the target object to which the fourth target point belongs is greater than the eleventh threshold value is judged. If the position variance of the target object to which the fourth target point belongs is greater than the eleventh threshold value, the confidence that the target object to which the fourth target point belongs is not an uninterested target is set to 0.8. If the position variance of the target object to which the fourth target point belongs is less than the eleventh threshold value, the confidence that the target object to which the fourth target point belongs is not an uninterested target is set to 0.2. Then the confidence that the target object to which the fourth target point belongs is not an uninterested target calculated in multiple frames of signals is fused by using the principle of evidence theory to obtain the probability that the target object to which the fourth target point belongs is not an uninterested target. The distance threshold value, the speed threshold value and the variance threshold value are all preset according to experience, which are not limited in the present application.

[0169] Based on the above scheme, the radar system can remove the influence of the uninterested target in the track by the method of grid map to calculate the uninterested probability, so that the generation of the track is more accurate.

[0170] The first target point may also include a second-reflected first target point. The second-reflected first target point is a kind of ghost point caused by multipath effect. After the signal is transmitted to the target, the signal is reflected to other objects, and the receiving antenna receives the signal reflected at the second-reflected first target point on the other objects. The radar system will mistakenly consider the first ghost point as a target point.

[0171] Referring to Figure 8 A second-reflected first target point diagram is provided for the embodiment of the present application. Specifically, the second-reflected first target point can be judged and the second-reflected probability can be calculated according to the following characteristics: the difference between the azimuth angle of the ghost point caused by the second reflection and the azimuth angle of the target point is less than a certain threshold value, the difference between the Doppler velocity of the ghost point and the Doppler velocity of the target point is less than a certain threshold value or the Doppler velocity of the ghost point is twice the Doppler velocity of the target point, and the distance of the ghost point is twice the distance of the target point. The threshold values are all preset according to experience, which are not limited in the present application.

[0172] For the ghost point caused by multipath effect which is difficult to be directly detected, the first ghost point probability needs to be calculated. Referring to Figure 9 A multipath target point reflection diagram is provided for the embodiment of the present application. The ghost point B and the ghost point C in the diagram are two cases of ghost points.

[0173] The path that the ghost point B can generate is S->A->D->A->S and S->D->A->S. It is judged whether the ghost point B satisfies the following characteristics compared with the real target point A:

[0174] Characteristic 1, the radial distance of the ghost point B is greater than the radial distance of the target point A.

[0175] Characteristic 2, the Doppler velocity of the ghost point B is between the right Doppler velocity and the left Doppler velocity, wherein the left Doppler velocity satisfies Vdoppler_Left = 0.5*(Velocity_A + dopper_A) Velocity_A, and the right Doppler velocity satisfies Vdoppler_Right = 0.5*(-Velocity_A + dopper_A), wherein Velocity_A is the speed of the target point A, and dopper_A is the Doppler velocity of the target point A.

[0176] Characteristic 3, the existence time of the ghost point B is less than the existence time of the real target point A.

[0177] Characteristic 4, the position of the ghost point B and the position of the real target point A are less than a certain threshold, and the heading angle of the ghost point B and the heading angle of the real target point A are less than a certain threshold, wherein the threshold is obtained according to experience, which is not limited in the present application.

[0178] Characteristic 5, the correlation duty cycle of the ghost point B is less than a certain threshold, wherein the correlation duty cycle is the ratio of the correlation time to the existence time, and the correlation time represents the time of judging that the first ghost point is the target point to which the first target belongs.

[0179] For the ghost point B satisfying the above characteristics, the second ghost point probability is calculated. Specifically, the confidence degree of B being a ghost point is set to 1, and the confidence degree of B being a ghost point is set to 0.3 if all the above characteristics cannot be satisfied. Then, the first ghost point probability is obtained by low-pass filtering combined with the results of the adjacent two frames of signals.

[0180] The path that the ghost point C can generate is S->A->D->S and S->D->A->D->S. It is judged whether the ghost point C satisfies the following characteristics compared with the real target point A:

[0181] Characteristic A, the radial distance of the ghost point C is greater than the radial distance of the target point A.

[0182] Feature B, the Doppler velocity of the ghost point C is between the right Doppler velocity and the left Doppler velocity, wherein the left Doppler velocity satisfies Vdoppler_Left = 0.5*(Velocity_A + dopper_A) Velocity_A, and the right Doppler velocity satisfies Vdoppler_Right = 0.5*(-Velocity_A + dopper_A), wherein Velocity_A is the velocity of the target point A, and dopper_A is the Doppler velocity of the target point A.

[0183] Feature C, the position of the ghost point C is greater than a certain threshold value from the position of the real target point A, and the heading angle of the ghost point C is greater than a certain threshold value from the heading angle of the real target point A.

[0184] Feature D, the Doppler velocity of the ghost point C is between Vdoppler_1 and Vdoppler_2, wherein Vdoppler_1 is 0.5*(MirrorDoppler + dopper_A - 2), Vdoppler_2 is 0.5*(MirrorDoppler + dopper_A + 2), dopper_A is the Doppler velocity of the target point A, and MirrorDoppler is calculated according to the path possibly generated by the ghost point C and the corresponding geometric model.

[0185] Feature E, the correlation duty cycle of the ghost point C is less than a certain threshold value.

[0186] For the ghost point C satisfying the above features, a second ghost point probability is calculated. Specifically, the confidence of C being a ghost point is set to 1, and the confidence of C being a ghost point is set to 0.3 if all the above features cannot be satisfied. Then, the result of the current frame is low-pass filtered with the result of the last frame to obtain a first ghost point probability.

[0187] Optionally, a multipath probability is calculated according to the probability of B being a ghost point and the probability of C being a ghost point.

[0188] Based on the above scheme, the influence of the ghost point caused by the multipath effect in the track can be reduced, and the track of the target object generated by the radar system is more accurate.

[0189] Optionally, the target detection probability and the false alarm probability can also be combined when updating the track. The target detection probability represents the probability that the target will be detected by the radar system, which is determined by the performance of the radar system. The false alarm probability represents the probability that the radar system will judge that there is a target due to the universal existence and fluctuation of noise in the radar detection process.

[0190] The target detection probability can be calculated according to the azimuth angle of the target point, satisfying the following formula.

[0191] P = exp(-1 / cos(f))

[0192] wherein f is an azimuth angle of the target point.

[0193] The false alarm probability is calculated according to the probability of invalid points caused by large targets and the secondary reflection probability. For example, if the secondary reflection probability of a first target point is high, the false alarm probability of the first target point can be 0.95, and if the secondary reflection probability of a first target point is low and the probability of invalid points caused by large targets is also low, the false alarm probability of the first target point can be 0.05.

[0194] Specifically, the invalid points caused by large targets refer to the case that due to the target being too large, invalid points are generated near the target when the target approaches the radar. The probability of invalid points caused by large targets is calculated according to the following process: searching for the SNR of all first target points near large targets, then comparing the SNR of the first target point with the SNR threshold of the first target point, if the SNR of the first target point is less than the SNR threshold of the first target point, the first target point is an invalid point, and the probability of the first target point being an invalid point is calculated. The SNR threshold of each first target point is preset according to the sidelobe attenuation coefficient of the window function. It should be understood that the definition of near and large target is obtained according to experience. For example, near can be within 20m, and large target can be a truck or other large target.

[0195] Based on the same concept of the above method, see Figure 10 A track generation device 1000 is provided for the embodiments of the present application. The device 1000 can perform each step in the above method, and to avoid repetition, it will not be described in detail here. The device 1000 includes a transceiver unit 1001 and a processing unit 1002.

[0196] In one scenario:

[0197] The transceiver unit 1001 is configured to transmit and receive signals.

[0198] The processing unit 1002 is configured to perform the following operations:

[0199] Controlling each transmitting antenna in an antenna array of a radar to transmit a signal simultaneously, and attaching a phase to the signal transmitted by each transmitting antenna; performing fast Fourier transform on each signal received by each receiving antenna in the antenna array to obtain an RV map of the each signal; determining a plurality of first target points in the RV map; the plurality of first target points are points with signal strength greater than a first threshold; determining a first track of a first target object according to the plurality of first target points; the first track is used to represent the trajectory of the first target object.

[0200] In a possible implementation, the processing unit 1002 is configured to determine a track of the first target object according to the plurality of first target points, and is specifically configured to: associate the plurality of first target points in two adjacent frames of signals in a time domain respectively to obtain a second track of a plurality of second target points; the plurality of second target points are part or all of the plurality of first target points; associate the plurality of second target points in a single frame of signal to obtain a first target object composed of the plurality of second target points; and combine the second tracks of the plurality of second target points to obtain the first track of the first target object.

[0201] In a possible implementation, the processing unit 1002 is configured to associate the plurality of first target points in two adjacent frames of signals in a time domain respectively to obtain a second track of a plurality of second target points, and is specifically configured to: associate the first target points satisfying one or more of the following conditions in two adjacent frames of signals in a time domain to obtain the second track of the plurality of second target points: an absolute value of a difference of distances of the first target points in two adjacent frames of signals in a time domain is less than or equal to a second threshold; if a distance of the first target point in a later frame is greater than a distance of the first target point in an earlier frame, a ratio of a difference of distances of the first target points to a frame period of the signal is greater than a third threshold; if the distance of the first target point in the later frame is less than or equal to the distance of the first target point in the earlier frame, the ratio of the difference of distances of the first target points to the frame period of the signal is less than or equal to a fourth threshold; and an absolute value of a difference of de-blurring Doppler velocities of the first target points in two adjacent frames of signals in a time domain is less than or equal to a fifth threshold.

[0202] In a possible implementation, the processing unit 1002 is further configured to: associate the plurality of first target points in a single frame of signal to obtain a plurality of second target objects composed of the plurality of first target points; the first target object is one of the plurality of second target objects; if there is no second track of a third target point in two adjacent frames of signals in a time domain, associate the third target point with a second target object to which the third target point belongs; and use position information of the second target object to which the third target point belongs in the single frame of signal as starting position information of the second target object to which the third target point belongs.

[0203] In a possible implementation, the processing unit 1002 is further configured to: determine azimuth angles and Doppler velocities of the plurality of first target points in a single frame of signal; determine first ghost point probabilities of the plurality of first target points according to the azimuth angles and the Doppler velocities of the plurality of first target points in two adjacent frames of signals in a time domain; and update the first track according to the first ghost point probabilities.

[0204] In a possible implementation, the processing unit 1002 is configured to determine, in a single frame of signals, a second ghost point probability of the plurality of first target points according to the azimuth angle and the Doppler velocity of the plurality of first target points, and determine, in two frames of signals adjacent in time domain, the first ghost point probability of the plurality of first target points according to the second ghost point probability of the plurality of first target points in a single frame of signals, wherein the second ghost point probability is determined according to one or more of the following: a radial distance from the radar and a radial distance from the radar to the first target object; or a difference between the Doppler velocity and a Doppler velocity of the first target object; or an existing time length; or a distance from the first target object in a single frame of signals; or a heading angle; or an associated duty cycle, the associated duty cycle being a ratio of an associated time to the existing time length, the associated time representing a time at which the first ghost point is determined to be a target point to which the first target object belongs.

[0205] In a possible implementation, the processing unit 1002 is further configured to determine, in the first track, a secondarily reflected first target point, and update the first track according to the secondarily reflected first target point.

[0206] In a possible implementation, the processing unit 1002 is further configured to perform, for any fourth target point in a plurality of fourth target points included in the first track, the plurality of fourth target points being part or all of the plurality of first target points, the following operations: determining, in a single frame of signals, a search space of the any fourth target point according to an azimuth angle standard deviation of the any fourth target point and a distance standard deviation of the any fourth target point in the RV map, the deambiguation Doppler velocity of the fourth target point being less than or equal to a sixth threshold value; determining, in the single frame of signals, an occupied probability of each grid in the search space; performing, in two frames of signals adjacent in time domain, filtering fusion on each grid by using a principle of evidence theory, and determining, for each grid, that the grid is in an occupied state when the occupied probability of the grid exceeds a seventh threshold value; and determining, according to the occupied state, an uninterested probability of a target object to which the fourth target point belongs, the uninterested probability representing a probability that the target object to which the fourth target point belongs is an uninterested target.

[0207] In a possible implementation, before the processing unit 1002 performs the following operation on any fourth target point in the plurality of fourth target points contained in the first track, the processing unit 1002 is further configured to: determine, in the single frame signal, that a distance between a position of the target object to which the plurality of fourth target points belong and a starting position of the target object to which the plurality of fourth target points belong is less than or equal to an eighth threshold value; determine, in the single frame signal, that an absolute velocity of the target object to which the plurality of fourth target points belong is less than or equal to a ninth threshold value; determine that the target object to which the plurality of fourth target points belong exists in a grid greater than or equal to a tenth threshold value in a number of times in the first frame signal; calculate a position variance of the target object to which the plurality of fourth target points belong according to positions of the target object to which the plurality of fourth target points belong in each frame signal; and determine that the position variance of the target object to which the plurality of fourth target points belong is less than or equal to an eleventh threshold value.

[0208] In a possible implementation, the processing unit 1002 is configured to determine, in the RV map, the plurality of first target points by: performing peak detection and ordered constant false alarm rate (OS-CFAR) detection on each point in the RV map in a distance dimension to obtain a first detection result; performing OS-CFAR detection on the point whose first detection result is greater than a twelfth threshold value in a Doppler velocity dimension to obtain a second detection result; and determining the point whose second detection result is greater than or equal to a thirteenth threshold value as a first target point.

[0209] In a possible implementation, the transceiving unit 1001 is configured to: the antenna array is composed of three transmitting antennas and four receiving antennas, the three transmitting antennas are arranged at equal intervals with an interval of 1Lambda, the lowest part of the transmitting antenna in the middle and the lowest part of the two transmitting antennas on both sides have an interval of 0.5Lambda, the rightmost transmitting antenna is horizontally spaced apart from the leftmost transmitting antenna by 2Lambda, and is vertically spaced apart from the leftmost transmitting antenna by 0Lambda, and the four receiving antennas are arranged at equal intervals with an interval of 0.5Lambda and a vertical interval of 0Lambda.

[0210] With reference to Figure 11 The embodiments of the present application further provide an electronic device, which includes a processor 1101 and a transceiver 1102. The transceiver 1102 is configured to transceive signals, and the processor 1101 performs the operation steps of the method in any possible implementation of the above method. The processor 1101 can be configured to perform the operations of the processing unit 1002, and the transceiver 1102 can be configured to perform the operations of the transceiving unit 1001.

[0211] The embodiments of the present application further provide a computer storage medium, which stores a computer program, and the program is executed by a processor to implement the steps of any method described above.

[0212] Those skilled in the art can understand that all or part of the steps of the above-mentioned method embodiments can be completed by program instruction related hardware. The foregoing program can be stored in a computer readable storage medium. The program executes the steps of the above-mentioned method embodiments when executed; and the foregoing storage medium includes various storage media that can store program codes, such as ROM, RAM, magnetic disk or optical disk.

[0213] Although the specific embodiments of the present application are described above, those skilled in the art should understand that these are only illustrative, and the protection scope of the present application is defined by the appended claims. Those skilled in the art can make various changes or modifications to the embodiments without departing from the principles and essence of the present application, and these changes and modifications all fall within the protection scope of the present application. Although the preferred embodiments of the present application have been described, those skilled in the art can make additional changes and modifications to the embodiments once they know the basic creative concept. Therefore, the appended claims are intended to be interpreted as including the preferred embodiments and all changes and modifications falling within the scope of the present application.

[0214] Obviously, those skilled in the art can make various modifications and variations to the present application without departing from the scope of the present application. Thus, the present application also intends to include these modifications and variations, provided that they fall within the scope of the claims of the present application and their equivalent technologies.

Claims

1. A method of generating a flight path, characterized by, The method comprises the following steps: controlling each transmitting antenna in an antenna array of a radar to transmit a signal simultaneously, and adding a phase to the signal transmitted by each transmitting antenna; the antenna array comprises three transmitting antennas, each frame signal comprises three sub-frames, and the angles pointed to by the sub-frames transmitted by the three transmitting antennas are different; performing fast Fourier transform on each signal received by each receiving antenna in the antenna array to obtain an RV graph of the signal; determining a plurality of first target points in the RV graph; the plurality of first target points are points with a signal strength greater than a first threshold value; in two frame signals adjacent in a time domain, determining a first target point in the plurality of first target points that satisfies an association condition as a second target point, and respectively associating a plurality of determined second target points to obtain second tracks of the plurality of second target points; if the second tracks exist, associating the plurality of second target points in a single frame signal to obtain a first target object composed of the plurality of second target points, and merging the second tracks of the plurality of second target points to obtain a first track of the first target object; the first track is used to represent a trajectory of the first target object; if the second tracks do not exist, associating the plurality of first target points in a single frame signal to obtain a plurality of second target objects composed of the plurality of first target points, the first target object being one of the plurality of second target objects; associating a third target point with a second target object to which the third target point belongs, and taking position information of the second target object to which the third target point belongs in the single frame signal as starting position information of a track of the second target object to which the third target point belongs; the third target point is a first target point that does not satisfy the association condition in two frame signals adjacent in the time domain; wherein the position information of the second target object is determined according to the following method: for each third target point, respectively performing the following steps: determining a longitudinal velocity difference value, a transverse velocity difference value, a 45-degree oblique direction velocity difference value and a negative 45-degree oblique direction velocity difference value of the third target point according to an included angle of the third target point relative to a coordinate system in a previous frame and a current frame; and determining a heading angle and a velocity value of the third target point according to position information of the third target point and a numerical relationship between the velocity difference values of the third target point; performing weighted clustering on all third target points to obtain the position information of the second target object.

2. The method of claim 1, wherein, The step of associating the plurality of first target points in two frame signals adjacent in the time domain to obtain second tracks of a plurality of second target points comprises the following steps: in two frame signals adjacent in the time domain, associating the first target points that satisfy one or more of the following conditions to obtain the second tracks of the plurality of second target points: an absolute value of a difference value of distances of the first target points in the two frame signals adjacent in the time domain is less than or equal to a second threshold value; In two adjacent frames of signals in time domain, if the distance of the first target point in the latter frame is greater than the distance of the first target point in the former frame, the ratio of the difference value of the distance of the first target point to the frame period of the signals is greater than a third threshold value; if the distance of the first target point in the latter frame is less than or equal to the distance of the first target point in the former frame, the ratio of the difference value of the distance of the first target point to the frame period of the signals is less than or equal to a fourth threshold value. In two adjacent frames of signals in time domain, the absolute value of the difference of the de-ambiguity Doppler velocity of the first target point is less than or equal to a fifth threshold value.

3. The method according to claim 1 or 2, characterized in that, Further comprising: determining the azimuth angle and the Doppler velocity of the plurality of first target points in a single frame of signals; determining the first ghost point probability of the plurality of first target points according to the azimuth angle and the Doppler velocity of the plurality of first target points in two adjacent frames of signals in time domain; updating the first track according to the first ghost point probability.

4. The method of claim 3, wherein, The determining the first ghost point probability of the plurality of first target points according to the azimuth angle and the Doppler velocity of the plurality of first target points in two adjacent frames of signals in time domain comprises: determining a second ghost point probability according to the azimuth angle and the Doppler velocity of the plurality of first target points in a single frame of signals; determining the first ghost point probability of the plurality of first target points in two adjacent frames of signals in time domain according to the second ghost point probability of the plurality of first target points in a single frame of signals; wherein the second ghost point probability is determined according to one or more of the following: the radial distance of the radar and the radial distance of the first target from the radar; or the difference of the Doppler velocity and the Doppler velocity of the first target; or the existing time length; or the distance of the first target in a single frame of signals; or the heading angle; or the associated duty cycle; the associated duty cycle is the ratio of the associated time to the existing time length, and the associated time represents the time of judging the first ghost point as the target point to which the first target belongs.

5. The method of claim 1, wherein, Further comprising: determining a first target point of secondary reflection in the first track; updating the first track according to the first target point of secondary reflection.

6. The method according to any of claims 1-2 and 4-5, characterized in that, Further comprising: performing the following operations on any fourth target point in the plurality of fourth target points contained in the first track, the plurality of fourth target points being part or all of the plurality of first target points: determining the search space of the any fourth target point according to the azimuth angle standard deviation of the any fourth target point and the distance standard deviation of the any fourth target point in the RV graph in a single frame of signals, the de-ambiguity Doppler velocity of the fourth target point being less than or equal to a sixth threshold value; determining the occupancy probability of each grid in the search space in the single frame of signals; in two adjacent frames of signals in time domain, filtering and fusing each grid using the principle of evidence theory, and for each grid, when the occupancy probability of the grid exceeds a seventh threshold value, determining that the grid is in an occupied state; determining the uninterested probability of the target to which the fourth target point belongs according to the occupied state; the uninterested probability represents the probability that the target to which the fourth target point belongs is an uninterested target.

7. The method of claim 6, wherein, Before the operation of performing the following operation on any fourth target point in the first track containing a plurality of fourth target points, further comprising: Determining, in the single frame signal, that the distance between the position of the target object to which the plurality of fourth target points belong and the starting position of the target object to which the plurality of fourth target points belong is less than or equal to an eighth threshold value; Determining, in the single frame signal, that the absolute speed of the target object to which the plurality of fourth target points belong is less than or equal to a ninth threshold value; Determining that the target object to which the plurality of fourth target points belong exists in a grid with a number of times of occupation greater than or equal to a tenth threshold value in the first frame signal; Calculating the position variance of the target object to which the plurality of fourth target points belong according to the position of the target object to which the plurality of fourth target points belong in each frame signal; and determining that the position variance of the target object to which the plurality of fourth target points belong is less than or equal to an eleventh threshold value.

8. The method of claim 1, wherein, The operation of determining a plurality of first target points in the RV graph comprises: Performing peak detection and ordered constant false alarm rate (OS-CFAR) on each point in the RV graph in the distance dimension to obtain a first detection result; Performing OS-CFAR on the points whose first detection result is greater than a twelfth threshold value in the Doppler velocity dimension to obtain a second detection result; Determining the points whose second detection result is greater than or equal to a thirteenth threshold value as the first target points.

9. The method of claim 1, wherein, Comprise: The antenna array further comprises four receiving antennas, the three transmitting antennas are arranged at equal intervals with a spacing of 1Lambda, the lowest part of the middle transmitting antenna and the lowest part of the two transmitting antennas on both sides have a spacing of 0.5Lambda, the rightmost transmitting antenna is horizontally spaced from the leftmost transmitting antenna by 2Lambda, and vertically spaced by 0Lambda, and the four receiving antennas are arranged at equal intervals with a spacing of 0.5Lambda and a vertical spacing of 0Lambda.

10. A track generation device, characterized by Comprise: The transceiver unit is configured to transmit and receive signals. The processing unit is configured to perform the following operations: Controlling each transmitting antenna in the antenna array of the radar to transmit a signal simultaneously, and attaching a phase to the signal transmitted by each transmitting antenna; the antenna array comprises three transmitting antennas, each frame signal comprises three sub-frames, and the angles pointed to by the sub-frames transmitted by the three transmitting antennas are different; Performing fast Fourier transform on each signal received by each receiving antenna in the antenna array to obtain an RV graph of the signal; Determining a plurality of first target points in the RV graph; the plurality of first target points are points with a signal strength greater than a first threshold value; In two adjacent frame signals in the time domain, determining first target points in the plurality of first target points that satisfy an association condition as second target points, and respectively associating the determined plurality of second target points to obtain second tracks of the plurality of second target points; If the second tracks exist, associating the plurality of second target points in a single frame signal to obtain a first target object composed of the plurality of second target points, and merging the second tracks of the plurality of second target points to obtain a first track of the first target object; The first track is used to represent the trajectory of the first target object. If the second track does not exist, the processing unit is configured to: in a single frame signal, associate the plurality of first target points to obtain a plurality of second target objects composed of the plurality of first target points, the first target object being one of the plurality of second target objects; associate a third target point with a second target object to which the third target point belongs, and take position information of the second target object to which the third target point belongs in the single frame signal as starting position information of a track of the second target object to which the third target point belongs; the third target point being a first target point that does not satisfy the association condition in two adjacent frames in the time domain; The position information of the second target object is determined according to the following method: For each third target point, the processing unit is configured to: determine a longitudinal velocity difference value, a transverse velocity difference value, a 45-degree oblique direction velocity difference value and a negative 45-degree oblique direction velocity difference value of the third target point according to an included angle of the third target point relative to the coordinate system in the last frame and the current frame; and determine a heading angle and a speed value of the third target point according to the position information of the third target point and a numerical relationship between the velocity difference values of the third target point. The processing unit is configured to: in two adjacent frames in the time domain, associate the plurality of first target points to obtain a plurality of second target points; and perform weighted clustering on all third target points to obtain the position information of the second target object.

11. The apparatus of claim 10, wherein, The processing unit is configured to: in two adjacent frames in the time domain, associate the plurality of first target points to obtain a plurality of second target points; and perform weighted clustering on all third target points to obtain the position information of the second target object. The processing unit is configured to: in two adjacent frames in the time domain, associate the plurality of first target points to obtain a plurality of second target points; and perform weighted clustering on all third target points to obtain the position information of the second target object. The processing unit is configured to: in two adjacent frames in the time domain, associate the plurality of first target points to obtain a plurality of second target points; and perform weighted clustering on all third target points to obtain the position information of the second target object. The processing unit is configured to: in two adjacent frames in the time domain, associate the plurality of first target points to obtain a plurality of second target points; and perform weighted clustering on all third target points to obtain the position information of the second target object. The processing unit is configured to: in a single frame signal, determine an azimuth angle and a Doppler velocity of the plurality of first target points; in two adjacent frames in the time domain, determine a first ghost point probability of the plurality of first target points according to the azimuth angle and the Doppler velocity of the plurality of first target points; and update the first track according to the first ghost point probability.

12. The apparatus of claim 10 or 11, wherein, The processing unit is configured to: in a single frame signal, determine an azimuth angle and a Doppler velocity of the plurality of first target points; in two adjacent frames in the time domain, determine a first ghost point probability of the plurality of first target points according to the azimuth angle and the Doppler velocity of the plurality of first target points; and update the first track according to the first ghost point probability. The processing unit is configured to: in a single frame signal, determine an azimuth angle and a Doppler velocity of the plurality of first target points; in two adjacent frames in the time domain, determine a first ghost point probability of the plurality of first target points according to the azimuth angle and the Doppler velocity of the plurality of first target points; and update the first track according to the first ghost point probability. ​ ​ 13. The apparatus of claim 12, wherein, ​ ​ According to the second ghost point probability of the plurality of first target points in a single frame signal, a first ghost point probability of the plurality of first target points in two frames of signals adjacent in time domain is determined. The second ghost point probability is determined according to one or more of the following: A radial distance from the radar and a radial distance from the radar to the first target; or A difference between a Doppler velocity and a Doppler velocity of the first target; or An existing time length; or A distance from the first target in a single frame signal; or A heading angle; or An associated duty cycle, which is a ratio of an associated time to the existing time length, the associated time representing a time of judging the first ghost point as a target point to which the first target belongs.

14. The apparatus of claim 10, wherein, The processing unit is further configured to: Determine a second-reflected first target point in the first track; and Update the first track according to the second-reflected first target point.

15. The apparatus of any of claims 10-11 and 13-14, wherein, The processing unit is further configured to: Perform the following operations on any fourth target point in a plurality of fourth target points included in the first track, the plurality of fourth target points being part or all of the plurality of first target points: Determine a search space of the any fourth target point in a single frame signal according to an azimuth angle standard deviation of the any fourth target point and a distance standard deviation of the any fourth target point in the RV graph, a deambiguating Doppler velocity of the fourth target point being less than or equal to a sixth threshold value; Determine an occupied probability of each grid in the search space in the single frame signal; Filter and fuse each grid in time domain using a principle of evidence theory, and determine that the grid is in an occupied state when the occupied probability of the grid exceeds a seventh threshold value. Determine an uninterested probability of a target to which the fourth target point belongs according to the occupied state, the uninterested probability representing a probability that the target to which the fourth target point belongs is an uninterested target.

16. The apparatus of claim 15, wherein, Before performing the following operations on any fourth target point in a plurality of fourth target points included in the first track, the processing unit is further configured to: Determine that a distance between a position of a target to which the plurality of fourth target points belong and a starting position of the target to which the plurality of fourth target points belong is less than or equal to an eighth threshold value in the single frame signal; Determine that an absolute velocity of the target to which the plurality of fourth target points belong is less than or equal to a ninth threshold value in the single frame signal; Determine that the target to which the plurality of fourth target points belong exists in a grid whose occupied times are greater than or equal to a tenth threshold value in a first frame signal; Calculate a position variance of the target to which the plurality of fourth target points belong according to positions of the target to which the plurality of fourth target points belong in each frame signal; Determine that the position variance of the target to which the plurality of fourth target points belong is less than or equal to an eleventh threshold value.

17. The apparatus of claim 10, wherein, The processing unit is configured to: Perform peak detection and ordered constant false alarm rate (OS-CFAR) detection on each point in the RV graph in a distance dimension to obtain a first detection result. performing OS-CFAR on the first detection result of the point greater than the twelfth threshold value in the Doppler velocity dimension to obtain a second detection result; determining a point with the second detection result greater than or equal to a thirteenth threshold value as a first target point.

18. The apparatus of claim 10, wherein, The transceiver unit is configured to: The antenna array further includes four receiving antennas, the three transmitting antennas are arranged at equal intervals with an interval of 1 Lambda, the interval between the lowest part of the middle transmitting antenna and the lowest part of the two transmitting antennas on both sides is 0.5 Lambda, the horizontal interval between the rightmost transmitting antenna and the leftmost transmitting antenna is 2 Lambda, and the vertical interval is 0 Lambda, and the four receiving antennas are arranged at equal intervals with an interval of 0.5 Lambda and a vertical interval of 0 Lambda.

19. An electronic device, comprising: comprising: a transceiver for receiving and transmitting signals; a processor connected with the transceiver, for processing the signals received and transmitted in the transceiver, and implementing the method in any one of claims 1 to 9 when processing the signals.

20. A computer readable storage medium, characterized in that, the computer readable storage medium stores computer instructions, when the computer instructions run on a computer, make the computer execute the method in any one of claims 1 to 9.

Images