Radar apparatus and signal processing method

Abstract

To provide a radar device capable of maintaining or realizing azimuth estimation of a target with high accuracy.SOLUTION: The memory unit 32 stores a correction matrix and section parameter information set for each correction section in accordance with the specifications of the protection member 3. The array correction unit 38 performs array correction on the array mode vector using the correction matrix to calculate a corrected array mode vector. An azimuth estimation processing part 35 estimates the azimuth of the target by using the radio wave transmitted / received by the array antenna and the corrected array mode vector. The direction error estimation processing unit 36 calculates the direction of the target from the relative speed of the target, and stores peak data obtained by frequency-analyzing the radio wave received when the direction of the target is calculated, in the bin of the correction section corresponding to the direction. The correction matrix calculating unit 37 calculates a correction matrix of a correction section in which peak data is stored in at least all bins of the correction section among a plurality of correction sections, and updates the correction matrix of the correction section.SELECTED DRAWING: Figure 18

Description

[Technical Field]

[0001] The present disclosure relates to a radar device and a signal processing method therefor. [Background technology]

[0002] Conventionally, radar devices are known that are mounted on vehicles and transmit and receive radio waves via the bumper from an array antenna installed between the bumper and the vehicle body to estimate the distance to a target outside the vehicle, the relative speed to the target, the target's direction, etc. In this type of radar device, the transmitted waves transmitted from the transmitting antenna and the incoming waves reflected by the target and arriving at the receiving antenna are affected by refraction and reflection by the bumper, which causes errors in the amplitude and phase of the received signal, which are direction-dependent, and mutual coupling errors between the antennas, resulting in a deterioration of direction estimation accuracy. One of the measures to maintain or achieve high direction estimation accuracy is a technology that corrects the above errors. The radar device described in Patent Document 1 handles errors that are direction-dependent (hereinafter referred to as "direction-dependent errors") using calibration data, i.e., a calibration matrix, stored in a memory unit before shipping from the factory. In this disclosure, the calibration data is referred to as correction data, and the calibration matrix is ​​referred to as a correction matrix. [Prior art documents] [Patent documents]

[0003] [Patent Document 1] Japanese Patent Publication No. 2020-003334 Summary of the Invention [Problem to be solved by the invention]

[0004] After detailed investigation, the inventors found that this azimuth-dependent error depends on the distance between the array antenna and the bumper (hereinafter referred to as the "radar-bumper distance"). Therefore, if the radar-bumper distance changes due to the application of a load to the bumper while the vehicle is in use or due to aging, the above error will change, and the corresponding correction matrix will also change. The radar device described in Patent Document 1 cannot cope with a change in error when the radar-bumper distance changes due to aging or other reasons while the vehicle is traveling.

[0005] In view of the above, an object of the present disclosure is to provide a radar device and a signal processing method that can maintain or achieve highly accurate target azimuth estimation. [Means for solving the problem]

[0006] According to one aspect of the present disclosure, a radar device is mounted on a moving body (2), transmits and receives radio waves from an array antenna via a protective member (3), and estimates the direction of a target present outside the moving body, a memory unit (32) for storing a correction matrix set for each correction section that divides the range of the viewing angle of the array antenna in accordance with the specifications of the protective member in order to perform array correction for correcting an array mode vector that indicates the phase of radio waves transmitted and received by the array antenna, and section parameter information including an angle step that divides the correction section; an array correction unit (38) that calculates a corrected array mode vector by performing array correction on the array mode vector using a correction matrix; an azimuth estimation processing unit (35) for estimating the azimuth of a target using radio waves transmitted and received by the array antenna and a corrected array mode vector; an azimuth error estimation processing unit (36) that calculates the azimuth of the target from the relative speed between the moving body and the target while the moving body is moving, calculates an error from the azimuth of the target estimated by the azimuth estimation processing unit, and stores peak data obtained by frequency analysis of the radio waves received when calculating the azimuth of the target in a bin corresponding to the azimuth of the target calculated from the relative speed, among a plurality of bins obtained by dividing a correction section into angle steps; and a correction matrix calculation unit (37) that calculates a correction matrix for a correction interval in which peak data is stored in at least all of its own bins among the plurality of correction intervals, and updates the correction matrix for that correction interval.

[0007] According to this, even if the radar-bumper distance changes due to load application to the protective member during use of the mobile object or aging, the correction matrix calculation unit updates the correction matrix of the correction section accordingly. Therefore, the array correction unit can calculate the corrected array mode vector using the updated correction matrix. Therefore, the radar device can maintain or achieve high accuracy in target azimuth estimation by the azimuth estimation processing unit.

[0008] According to another aspect of the present disclosure, a signal processing method for a radar device mounted on a moving object (2), transmitting and receiving radio waves from an array antenna (3) via a protective member, and estimating a direction of a target, includes: storing, in a memory unit (S15), a correction matrix set for each correction section obtained by dividing the range of the viewing angle of the array antenna in accordance with the specifications of the protective member in order to perform array correction for correcting an array mode vector indicating the phase of radio waves transmitted and received by the array antenna, and section parameter information including an angle step for dividing the correction section; Calculating a corrected array mode vector by performing array correction using a correction matrix on the array mode vector (S247); Estimating the azimuth of the target using radio waves transmitted and received by the array antenna and the corrected array mode vector (S230); Calculating the azimuth of the target from the relative speed between the moving body and the target while the moving body is moving (S242); Among a plurality of bins obtained by dividing the correction section by angle steps, peak data obtained by frequency analysis of the radio waves received when calculating the target's direction is stored in a bin corresponding to the target's direction calculated from the relative speed (S244); The method includes calculating a correction matrix for a correction interval in which peak data is stored in at least all of its own bins among the plurality of correction intervals, and updating the correction matrix for that correction interval (S246).

[0009] According to this method, even if the radar-bumper distance changes while the mobile object is in use, the correction matrix of the correction section is updated accordingly. Therefore, the corrected array mode vector is calculated using the updated correction matrix. Therefore, this signal processing method can maintain or achieve high-accuracy target azimuth estimation.

[0010] The reference symbols in parentheses attached to each component indicate an example of the correspondence between the component and the specific components described in the embodiments described below. [Brief explanation of the drawings]

[0011] [Figure 1] 1 is a schematic configuration diagram of a radar device according to a first embodiment. [Figure 2] FIG. 2 is a schematic configuration diagram of a control unit included in the radar device. [Figure 3] FIG. 10 is an explanatory diagram for explaining interval parameters. [Figure 4] FIG. 10 is an explanatory diagram illustrating a state in which peak data is stored in all bins of some correction sections among a plurality of correction sections. [Figure 5] FIG. 10 is an explanatory diagram for explaining a method for setting section parameters etc. after the radar device is mounted on a vehicle and before the vehicle is shipped. [Figure 6] FIG. 10 is an explanatory diagram for explaining a method for setting section parameters etc. in a radar device manufacturing factory or the like. [Figure 7] FIG. 10 is an explanatory diagram for explaining input of section parameter information corresponding to the specifications of a bumper from an external database to a memory unit of the radar device. [Figure 8] 10 is a flowchart for explaining a section parameter setting process. [Figure 9] FIG. 10 is an explanatory diagram illustrating a method for determining a correction matrix. [Figure 10] FIG. 2 is an explanatory diagram for explaining a receiving section of the radar device. [Figure 11] FIG. 10 is an explanatory diagram for explaining an outline of a direction estimation process. [Figure 12] 10 is a graph illustrating the error between the estimated direction and the actual direction when the section parameters are not set appropriately. [Figure 13] 10 is a graph showing a MUSIC spectrum after correction with a correction matrix when the section parameters are not set appropriately. [Figure 14] 10 is a graph illustrating the error between the estimated direction and the actual direction when the section parameters are appropriately set. [Figure 15] 10 is a graph showing a MUSIC spectrum after correction with a correction matrix when the section parameters are appropriately set. [Figure 16] 10 is a graph showing the relationship between the radar-bumper distance and the heading error characteristics. [Figure 17] 4 is a flowchart illustrating signal processing executed by a control unit of a radar device mounted on a vehicle while the vehicle is traveling. [Figure 18] 10 is a flowchart illustrating details of a direction error estimation process. [Figure 19] 10 is a flowchart for explaining details of array correction processing for a corresponding section. [Figure 20] FIG. 2 is an explanatory diagram for explaining a method for calculating the azimuth of a target from a relative velocity. [Figure 21] 10 is a graph showing the relationship between the value obtained by dividing the relative speed of a target by the speed of the vehicle itself and the azimuth of the target. [Figure 22] 10 is a graph showing the relationship between the value obtained by dividing the relative speed of a target by the speed of the vehicle itself and the azimuth of the target. [Figure 23] FIG. 11 is an explanatory diagram for explaining a state in which peak data is stored in all bins required for calculating a correction matrix for some correction intervals among a plurality of correction intervals in the second embodiment. [Figure 24] FIG. 11 is an explanatory diagram for explaining a state in which peak data is stored in all bins of some correction intervals among a plurality of correction intervals in the third embodiment. DETAILED DESCRIPTION OF THE INVENTION

[0012] Hereinafter, embodiments of the present disclosure will be described with reference to the drawings. In the following embodiments, identical or equivalent parts will be denoted by the same reference numerals, and description thereof will be omitted.

[0013] (First embodiment) A first embodiment will be described. The radar device of the first embodiment is mounted between a bumper and the vehicle body of a vehicle, and transmits and receives radio waves from an array antenna via the bumper to estimate the distance between the vehicle and a target outside the vehicle, the relative speed between the target and the vehicle, the direction of the target relative to the vehicle, etc. The vehicle is an example of a "moving body" as defined in the claims, and the bumper is an example of a "protective member" as defined in the claims.

[0014] The radar device employs the FMCW method as a signal modulation method. The radar device operates at a frequency band corresponding to millimeter waves (e.g., 76.5 GHz). Note that the operating frequency of the radio waves transmitted and received by the radar device is not limited to a frequency corresponding to millimeter waves, and may be a frequency other than millimeter waves.

[0015] As shown in Fig. 1, the radar device 1 includes a transmitter 10, a receiver 20, and a controller 30. The transmitter 10 has an oscillator 11 and a modulator 12, and transmits a transmission wave from a transmitter antenna 13. The receiver 20 receives, using a plurality of receiver antennas 21, reflected waves of the transmission wave reflected by targets. The receiver 20 processes the signal received by the receiver antenna 21 in an MMIC 22 to convert it into a received signal X of a desired form. Note that MMIC is an abbreviation for Monolithic Microwave Integrated Circuit.

[0016] The control unit 30 is an electronic control unit having a microcomputer equipped with a processor 31 and a memory unit 32. The memory unit 32 is composed of, for example, ROM, RAM, EEPROM, etc. The memory unit 32 is a non-transitory tangible recording medium. The various functions of the microcomputer are realized by the processor 31 executing computer programs stored in the memory unit 32. Note that some or all of the functions executed by the processor 31 may be realized by hardware circuits.

[0017] The control unit 30 uses a processor 31 to perform array correction, which corrects an array mode vector indicating a phase corresponding to the angle of arrival of radio waves transmitted and received by the array antenna, using a correction matrix. The control unit 30 also uses the processor 31 to estimate the azimuth of a target using a corrected array mode vector, which is the result of the array mode vector correction, and a received signal X corresponding to the radio waves transmitted and received by the array antenna. As shown in FIG. 3, the memory unit 32 stores correction matrices set for each correction interval, which divides the range of the field of view angle of the array antenna. The memory unit 32 also stores interval parameter information, including information such as the number of correction intervals, the interval width, and the angle step for dividing the correction interval into multiple regions. Each region obtained by dividing the correction interval by the angle step is referred to as a "bin." The interval parameters are set according to the bumper specifications (e.g., curvature, paint material, number of coats).

[0018] The radar device 1 of this embodiment is configured as a SIMO radar having one transmitting antenna 13. The radar device 1 may be configured as a MIMO radar or a MISO radar having multiple transmitting antennas 13. SIMO is an abbreviation for Single Input Multiple Output. MIMO is an abbreviation for Multiple Input Multiple Output. MISO is an abbreviation for Multiple Input Single Output. A MIMO radar having m transmitting antennas 13 and n receiving antennas 21 can be treated as an array antenna equivalent to a SIMO radar having m×n receiving antennas 21. A MISO radar having m transmitting antennas 13 and one receiving antenna 21 can be treated as an array antenna equivalent to a SIMO radar having m receiving antennas 21.

[0019] As shown in FIG. 2, the control unit 30 has functions such as an FFT processing unit 33, an FFT peak extraction processing unit 34, an orientation estimation processing unit 35, an orientation error estimation processing unit 36, a memory unit 32, a correction matrix calculation unit 37, and an array correction unit 38.

[0020] When the receiving antenna 21 receives a wave reflected by a target, the FFT processing unit 33 performs frequency analysis on the received signal X using FFT. The FFT peak extraction processing unit 34 extracts frequency peak data of the received signal X from the result of the frequency analysis performed by the FFT processing unit 33. The direction estimation processing unit 35 estimates the direction of the target using the corrected array mode vector and the received signal X. The direction error estimation processing unit 36 ​​calculates the direction of the target from the relative speed between the traveling vehicle and the target. The direction error estimation processing unit 36 ​​then calculates the error between the direction of the target calculated from the relative speed and the direction of the target estimated by the direction estimation processing unit 35.

[0021] As described above, the memory unit 32 stores section parameter information. The memory unit 32 is configured to be able to input section parameter information corresponding to the specifications of the bumper of the vehicle equipped with the radar device 1 from the external database 40. Of the multiple bins of the section parameters, peak data extracted by frequency analysis of the radio waves received when calculating the target direction is stored in a bin corresponding to the target direction calculated from the relative speed between the vehicle and the target. As shown in FIG. 4, peak data is gradually stored in each bin of the section parameters while the vehicle is traveling. In FIG. 4, bins in which peak data is stored are indicated by black circles. In the example shown in FIG. 4, peak data is stored in all bins of the correction section Q3.

[0022] The correction matrix calculation unit 37 shown in Fig. 2 calculates a correction matrix for a correction interval in which peak data is stored in all of its bins among multiple correction intervals. The correction matrix calculation unit 37 then updates the correction matrix for the correction interval for which the correction matrix was calculated. The array correction unit 38 performs array correction on the array mode vector using the correction matrix to obtain a corrected array mode vector. The corrected array mode vector is used by the azimuth estimation processing unit 35 to estimate the azimuth of the target.

[0023] Next, the "section parameter setting process, etc." executed by the control unit 30 will be described with reference to Figs. 5 to 11. The section parameter setting process, etc., is a calculation process for obtaining section parameter information and an initial correction matrix corresponding to the specifications of a predetermined bumper provided on a vehicle on which the radar device 1 is mounted. In the following description, the steps of the control process will simply be represented as "S."

[0024] As shown in Fig. 5, the section parameter setting process is performed, for example, after the radar device 1 is mounted on the vehicle 2 and before the vehicle is shipped. The radar device 1 is mounted between a bumper 3 and the body of the vehicle 2. In this section parameter setting process, for example, a turntable 4 that rotates the vehicle 2 on which the radar device 1 is mounted as indicated by arrow R, and a corner reflector 60 that serves as a target are used. Alternatively, as shown in Fig. 6, the section parameter setting process is performed, for example, in a manufacturing plant for the radar device 1, using the radar device 1, multiple types of bumpers 3, the turntable 4, and the corner reflector 60.

[0025] First, the process of setting section parameters etc., which is executed after the radar device 1 is mounted on the vehicle 2 and before the vehicle is shipped, will be described.

[0026] 8, in S10, the radar device 1 transmits a chirp signal from the transmitting antenna 13 at a predetermined transmission cycle, and acquires received signals X when the reflected waves from the corner reflectors 60 are received by the multiple receiving antennas 21 in a known direction. Specifically, the vehicle 2 is rotated together with the rotating platform 4 in predetermined angular steps within the range of the field of view of the radar device 1. Then, the radar device 1 stores the received signals X measured each time the vehicle 2 is rotated in the predetermined angular steps in the memory unit 32 as received data.

[0027] Next, in S11, the control unit 30 sets section parameters such as the number of sections, section width, angle step, etc. of the correction section. As a method of setting the section parameters, for example, a predetermined initial value is set the first time, and from the next time onwards, any one of the number of sections, section width, and angle step of the correction section is changed and set.

[0028] Next, in S12, the control unit 30 executes a calculation process to obtain a correction matrix Q for each correction interval. There are various methods for obtaining the correction matrix Q, which can be roughly divided into methods A and B shown in FIG. 9. Method A obtains the correction matrix Q by using a received signal Xmeans received from a known arrival angle or a signal subspace U obtained by eigenvalue decomposition of the received signal Xmeans. Sand the ideal array mode vector determined by the known arrival angle. Method B calculates the correction matrix Q using the noise subspace U obtained by eigenvalue decomposition of the received signal Xmeans received from the known arrival angle. N and an ideal array mode vector determined by a known angle of arrival. The received signal Xmeans is obtained by averaging the actual received signal X over a predetermined number of snapshots.

[0029] Next, in S13, the control unit 30 executes a direction estimation process for estimating the direction of the corner reflector 60.

[0030] For example, as shown in FIG. 10, the receiving unit 20 of the radar device 1 is a linear array having L receiving antennas 21 (element positions d 1、 d2, d L ) and the direction of the arriving wave is θ K is estimated. In this case, the received signal X at each receiving antenna 21 is not an ideal signal, but contains various array errors such as mutual coupling between elements, errors due to variations in the characteristics (amplitude, phase) of each antenna element, and errors due to the position of the antenna element. The received signal X containing such array errors can be modeled, for example, by equations F10 to F15 in FIG. 11. In equation F10, the received signal X is represented as "X(t)".

[0031] Here, T in formulas F10 to F13 and F15 is a transpose. "s(t)" in formula F12 is a signal vector in which each element represents the complex amplitude of the kth arriving wave. "n(t)" in formulas F11 and F13 is a noise vector in which the mean at each antenna element is "0" and the variance is noise consisting of noise components of thermal noise power. "a(θ K )” is the ideal array mode vector corresponding to the kth arriving wave, and “A” is “a(θ K )" as columns. The array mode vector is a vector that indicates the phase according to the angle of arrival of the arriving wave. "λ" is the wavelength.

[0032] Furthermore, "Q(θ)" in formulas F11 and F16 is a matrix of a generalized error model that includes any error, such as errors due to variations in mutual coupling between elements and the characteristics (amplitude, phase) of each antenna element, or errors due to element position, and is expressed by formula F16. Q(θ) is called an angle-dependent correction matrix.

[0033] When the received signal X is acquired from the receiving unit 20, the FFT processing unit 33 performs FFT processing. FFT is an abbreviation for Fast Fourier Transform. Next, the FFT peak extraction processing unit 34 performs FFT peak processing. Next, the array correction unit 38 performs array correction processing. The array correction processing is a process of correcting the ideal array mode vector to obtain a corrected array mode vector that takes into account various array errors, etc. In the array correction processing, the ideal array mode vector "a(θ)" is corrected using a correction matrix Q(θ) to obtain the corrected array mode vector "Q(θ)a(θ)".

[0034] Next, the direction estimation processor 35 estimates the direction using the corrected array mode vector and the received signal X. Specifically, the direction estimation processor 35 calculates the correlation matrix R shown in equation F17 of FIG. XX In addition, "T" in the formula F17 is not a transpose but the number of snapshots. Then, the direction estimation processing unit 35 calculates the correlation matrix R XX From the eigenvalue decomposition of N At this time, the array mode vector Q(θ)a(θ) corrected by the correction matrix and the noise subspace U N The relationship is Q(θk)a(θk)⊥ U N (k=1, 2,...k), and the MUSIC method shown in formula F19 makes use of this relationship. When θ=θk(k=1, 2,...k) in formula 19, the denominator of formula 19 approaches 0, resulting in a very large value. The direction can be estimated by detecting this peak value.

[0035] In this embodiment, the direction is estimated by the MUSIC method, but the direction estimation is not limited to the MUSIC method and may be realized by other methods such as DBF, Capon, and MODE.

[0036] Next, in S14, the control unit 30 determines whether the estimation error of the direction obtained by the direction estimation process is equal to or less than a predetermined threshold. In this determination process, the estimated direction obtained by the direction estimation process is compared with a known angle of arrival to determine the estimation error of the estimated direction. The control unit 30 then determines whether the estimation error of the estimated direction is equal to or less than a threshold. The threshold is set, for example, to the upper limit of the allowable range of direction error.

[0037] If the estimation error exceeds the threshold, the control unit 30 returns the process to S11, calculates the correction matrix Q using section parameters different from those used so far, and performs the direction estimation process using the correction matrix Q.

[0038] On the other hand, if the estimation error is equal to or less than the threshold, in S15 the control unit 30 outputs the section parameters for which the estimation error is equal to or less than the threshold as optimal section parameter information. The control unit 30 also outputs the correction matrix calculated in the correction matrix calculation process as an initial correction matrix. The section parameter information and initial correction matrix calculated in this way are stored in the memory unit 32 of the radar device 1 and are used when estimating the azimuth of the target.

[0039] Next, a process for setting section parameters, etc., which is executed, for example, at a manufacturing plant for the radar device 1 before the radar device 1 is mounted on a vehicle, will be described with reference to Figures 6 to 8. This method is effective when data cannot be acquired after the radar device 1 is mounted on the vehicle 2 and before the vehicle is shipped.

[0040] As shown in Fig. 6, this section parameter setting process is performed using a bumper 3, a radar device 1, a rotating table 4, and a target corner reflector 60. A plurality of types of bumpers 3 with different specifications (i.e., curvature, coating material, number of coatings) are prepared. The rotating table 4 is capable of rotating both the bumper 3 and the radar device 1 as shown by the arrow R.

[0041] 8, the radar device 1 transmits a chirp signal from the transmitting antenna 13 at a predetermined transmission cycle and acquires a received signal X from a known direction in which the corner reflector 60 is installed. Specifically, the bumper 3 and the radar device 1 are rotated together by the rotating table 4 in predetermined angle steps within the range of the field of view of the radar device 1. Then, the radar device 1 stores the received signal X measured each time the rotating table 4 is rotated in the predetermined angle steps in the memory unit 32 as received data.

[0042] The processes of S11 to S14 are the same as those described above. In S15, the control unit 30 outputs the optimum section parameter information and the initial correction matrix to the external database 40. In the section parameter setting process shown in Fig. 6, the optimum section parameter information and the initial correction matrix are set for multiple types of bumpers 3 with different specifications in terms of curvature, coating material, and number of coats, and the information is output to the external database 40.

[0043] The external database 40 stores a section parameter table as shown in Fig. 7. In the section parameter table, optimal section parameter information and initial correction matrices for each bumper specification set in the section parameter setting process are stored in each cell corresponding to the bumper curvature, coating material, and number of coats. That is, the section parameter information as shown in Fig. 3 is stored in each cell of the section parameter table. Note that the section parameter information stored in each cell of the section parameter table differs depending on the bumper specification.

[0044] Once the vehicle on which the radar device 1 is to be mounted has been determined and the specifications of the bumper of that vehicle have been determined, the optimal section parameter information and initial correction matrix corresponding to the bumper specifications are input from the external database 40 to the memory unit 32 of the radar device 1. For example, in FIG. 8 , if the specifications of the bumper of the vehicle on which the radar device 1 is mounted are curvature R1, paint material P1, and one coating cycle, the optimal section parameter information and initial correction matrix stored in the cell indicated by the dashed square are input to the memory unit 32 of the radar device 1.

[0045] However, when applying the section division correction technique as a means of correcting azimuth-dependent errors to bumper errors (i.e., errors caused by the bumper), the inventors' detailed study revealed that the section parameter information required to correct the bumper error varies depending on the bumper specifications. Therefore, if the same section parameters are set uniformly for all bumpers, the number of sections and azimuth steps will be insufficient to correct the bumper error, and the azimuth estimation accuracy will not improve. Conversely, if the number of sections and azimuth steps is too large, problems such as increased data collection convergence time and computational complexity will occur. Furthermore, the inventors' detailed study also revealed that the distance between the array antenna and the bumper (hereinafter referred to as the "radar-bumper distance") changes with bumper errors, but the section parameters do not change.

[0046] Figure 12 shows an example of an actual measurement in which the section parameters were not set appropriately. Specifically, as shown in Figure 12(a), the section parameters for the correction data were set to 6 sections and an angle step of 3 degrees. In other words, a relatively small amount of correction data was used when calculating the correction matrix. The evaluation data (i.e., correct data) was prepared with an angle step of 0.5 degrees. In this case, as shown in the graph in Figure 12(b), the influence of the direction-dependent error becomes large, and the effect of reducing the direction estimation error is small. Figure 12(c) is an enlarged portion of the graph in Figure 12(b). As shown in the graph in Figure 12(c), the estimated direction error is zero for directions that are multiples of 3 degrees. Therefore, if the evaluation data direction is evaluated at the same 3 degrees as the correction data direction, the estimated direction error will be mistakenly assumed to be small, even though it is actually large. As shown in the graph in Figure 13, the MUSIC spectrum after correction with the correction matrix shows that the peak values ​​of the spectral power for each direction are closer to directions that are multiples of 3 degrees. This causes the phenomenon shown in FIG. 12(c).

[0047] In contrast, Figure 14 shows an example of an actual measurement in which the section parameters were appropriately set. Specifically, as shown in Figure 14(a), the section parameters for the correction data were set to 12 sections and an angle step of 1 degree. In other words, a relatively large amount of correction data was used when calculating the correction matrix. The evaluation data was prepared with an angle step of 0.5 degrees. In this case, as shown in the graph in Figure 14(b), the influence of the direction-dependent error was reduced, and the direction estimation error was sufficiently reduced. Figure 14(c) is an enlarged view of a portion of the graph in Figure 14(b). As shown in the graph in Figure 14(c), the estimated direction error is zero or close to zero for all directions. Therefore, even if the evaluation data direction is evaluated using a value different from the correction data direction, the estimated direction error is small. As shown in the graph in Figure 15, the MUSIC spectrum after correction with the correction matrix shows that the peak value of the spectral power for each direction indicates the correct direction. Therefore, it is preferable to set the section parameters optimal for correcting the bumper error according to the bumper specifications.

[0048] Next, Fig. 16 shows the direction estimation error in each direction when the radar-bumper distance is set to 19 mm, 20.5 mm, and 22 mm.

[0049] As shown in FIG. 16, for radar-to-bumper distances of 19 mm, 20.5 mm, and 22 mm, the change in the azimuth estimation error is small near the azimuth 0° (specifically, approximately -10 to 10°), but the change in the azimuth estimation error is large as the azimuth moves away from near 0°. Although not shown, experiments conducted by the inventors have revealed that a similar trend is observed even when the radar-to-bumper distance is set to a distance other than 19 mm, 20.5 mm, or 22 mm. This indicates that the radar-to-bumper distance changes the bumper error, but does not change the section parameters. Therefore, the same section parameter information can be input to the memory unit 32 of the radar device 1, even if the radar-to-bumper distance changes while the vehicle is in use, as long as the bumper specifications remain unchanged.

[0050] Next, the signal processing that the control unit 30 of the radar device 1 mounted on the vehicle executes while the vehicle is traveling will be described with reference to FIGS.

[0051] As shown in the flowchart of FIG. 17, in S200, the radar device 1 transmits a chirp signal from the transmitting antenna 13 at a predetermined transmission period, and acquires from the receiving unit 20 a received signal X when a reflected wave from a target is received by the multiple receiving antennas 21.

[0052] When the reception signal X is acquired from the reception unit 20, the FFT processing unit 33 performs FFT processing in S210. Note that FFT is an abbreviation for Fast Fourier Transform.

[0053] Next, in S220, the FFT peak extraction processor 34 performs FFT peak processing. The control unit 30 performs frequency analysis of the received signal X using FFT to determine the distance to the target. The control unit 30 also performs FFT on each beat frequency component to determine the Doppler frequency, and determines the relative velocity of the target based on the Doppler frequency. The control unit 30 creates an RV map consisting of the distance and relative velocity determined by the FFT processing.

[0054] Next, in S230, the direction estimation processing unit 35 performs direction estimation processing to estimate the direction of the target. The direction estimation processing method is the same as the direction estimation processing method described in S13 of the section parameter setting processing.

[0055] Next, in S240, the heading error estimation processing unit 36 ​​performs heading error estimation processing. Details of the heading error estimation processing are shown in the flowchart of FIG.

[0056] As shown in Fig. 18, in S241, the control unit 30 determines whether the host vehicle speed is greater than a predetermined speed threshold. The speed threshold is set to a value that provides a sufficiently large gradient for the graphs in Figs. 21 and 22 that show the relationship between the target's azimuth and the value obtained by dividing the target's relative speed by the host vehicle speed, as referenced in S242 described below. If the host vehicle speed is greater than the predetermined speed threshold, the process proceeds to S242. If the host vehicle speed is less than the predetermined speed threshold, the process ends as it is assumed that effective learning cannot be expected.

[0057] Next, in S242, the control unit 30 creates a distribution of data consisting of the measured relative speed between the target and the vehicle and the target's direction. This process calculates the direction from the relative speed for target data obtained during driving, in order to treat targets in unknown directions as if they were in known directions.

[0058] 20, the control unit 30 detects stationary targets such as guardrails, signs, curbs, etc. While the vehicle 2 is traveling, a stationary target present in front of the vehicle 2 is referred to as "target A," and a stationary target present at a predetermined angle θ VER1The stationary target B is located at a predetermined angle θ VER2 The stationary target present in the direction of the radar device 1 is referred to as "target C." The relative speed of target A facing the radar device 1 is the same as the vehicle speed. On the other hand, the relative speed of target B facing the radar device 1 is cosθ VER1 The relative speed of the target C heading towards the radar device 1 is cosθ VER2 This becomes:

[0059] In Figures 21 and 22, the horizontal axis is the vertical angle θ VER 21 is a graph showing the range in which a stationary target exists when the vertical axis is qv / Cmv. Note that qv is the magnitude of the relative speed of the stationary target in the XZ plane, and Cmv is the magnitude of the host vehicle speed in the XZ plane. The XZ plane is a plane defined by the vehicle longitudinal direction X and the vertical direction Z. As shown in FIG. 21, if there is no axial misalignment in the radar device 1, the beam center direction coincides with the vehicle longitudinal direction, and there is no bumper 3, the stationary target will be plotted on a semicircle 100 indicated by a dashed line.

[0060] However, in reality, the radar device 1 has an axis misalignment, and there is a bumper 3 in front of the radar device 1. Therefore, as shown in FIG. 22, a stationary target is plotted on a curve 102 that is shifted from the dashed-dotted semicircle 101. The dashed-dotted semicircle 101 in FIG. 22 indicates the position where a stationary target is plotted when there is an axis misalignment in the radar device 1 and there is no bumper 3 in front of the radar device 1. In FIG. 22, the magnitude of the axis misalignment is indicated by αv. The radar device 1 collects and compiles statistics of multiple target data, determines a representative point taking into account the axis misalignment of the radar device 1 and the misalignment caused by the bumper 3, and determines the target's azimuth. This makes it possible to treat a target with an unknown azimuth as a known azimuth.

[0061] Next, in S243 of FIG. 18, the azimuth error estimation processing unit 36 ​​calculates the error between the target azimuth calculated from the relative speed of the target in S242 and the target azimuth estimated by the azimuth estimation processing unit 35 in S230.

[0062] Next, in S244, the control unit 30 stores FFT peak data obtained by frequency analysis of the radio waves received when calculating the target's bearing in the bearing bins of the correction section corresponding to the bearing calculated from the relative speed. Hereinafter, the correction section corresponding to the bearing calculated from the relative speed will be referred to as the "corresponding section," and the bearing bins of the correction section corresponding to the bearing calculated from the relative speed will be referred to as the "bearing bins of the corresponding section." As shown in FIG. 4, peak data is gradually stored in each bin of the section parameters while the vehicle is traveling. In FIG. 4, bins in which peak data is stored are indicated by black circles. In the example shown in FIG. 4, peak data is stored in all bins of the correction section Q3.

[0063] In S245 of FIG. 18, the control unit 30 determines whether there are any free bins for FFT peak data in the corresponding interval. In other words, the control unit 30 determines whether there is any correction interval in which peak data is stored in all of its own bins among the multiple correction intervals. If there are any free bins for FFT peak data in the corresponding interval, i.e., there is no correction interval in which peak data is stored in all of its own bins among the multiple correction intervals (S245; NO), the process ends. On the other hand, if there are any free bins for FFT peak data in the corresponding interval, i.e., there is a correction interval in which peak data is stored in all of its own bins among the multiple correction intervals (S245; YES), the process proceeds to S246.

[0064] In S246, the correction matrix calculation unit 37 calculates a correction matrix for a correction interval in which peak data is stored in all of its own bins among the multiple correction intervals, and updates the correction matrix for the correction interval for which the correction matrix was calculated. This allows the radar device 1 to update the correction matrix for the correction interval to accommodate changes in the radar-bumper distance due to load application to the bumper while the vehicle is in use, aging, etc.

[0065] In S247, the array correction unit 38 performs array correction processing for the correction section. Details of the array correction processing are shown in the flowchart of FIG.

[0066] As shown in FIG. 19 , in S2471, the array corrector 38 calculates an ideal array mode vector a(θ). Next, in S2472, the array corrector 38 corrects the ideal array mode vector a(θ) using a correction matrix Q to calculate a corrected array mode vector Q(θ)a(θ). The corrected array mode vector is used to estimate the target's azimuth in the azimuth estimation process in S230. As described above, if the radar-bumper distance changes while the vehicle is in use, the radar device 1 updates the correction matrix of the correction section accordingly. Therefore, the array corrector 38 can calculate the corrected array mode vector using the updated correction matrix. Therefore, the radar device 1 can maintain or achieve high-accuracy target azimuth estimation by the azimuth estimation processor 35.

[0067] The radar device 1 and its signal processing method according to the first embodiment described above have the following configurations and provide the following effects.

[0068] (1) The radar device 1 of the first embodiment includes a memory unit 32, an array correction unit 38, an azimuth estimation processing unit 35, an azimuth error estimation processing unit 36, and a correction matrix calculation unit 37. The memory unit 32 stores a correction matrix and section parameter information set for each correction section according to the bumper specifications. The array correction unit 38 calculates a corrected array mode vector by performing array correction on the array mode vector using the correction matrix. The azimuth estimation processing unit 35 estimates the azimuth of a target using radio waves transmitted and received by the array antenna and the corrected array mode vector. The azimuth error estimation processing unit 36 ​​calculates the azimuth of a target from the relative speed between the vehicle and the target while the vehicle is moving, and calculates the error from the azimuth of the target estimated by the azimuth estimation processing unit 35. The azimuth error estimation processing unit 36 ​​also stores peak data obtained by frequency analysis of the radio waves received when calculating the azimuth of the target for a bin in the correction section corresponding to the azimuth of the target calculated from the relative speed. The correction matrix calculation unit 37 calculates a correction matrix for a correction interval in which peak data is stored in at least all of its own bins among a plurality of correction intervals, and updates the correction matrix for that correction interval. According to this, even if the radar-bumper distance changes due to the application of a load to the bumper during use of the vehicle or due to aging, the correction matrix calculation unit 37 updates the correction matrix of the correction section accordingly. Therefore, the array correction unit 38 can calculate the corrected array mode vector using the updated correction matrix. Therefore, the radar device 1 can maintain or achieve high accuracy in target azimuth estimation by the azimuth estimation processing unit 35. Furthermore, this radar device 1 can improve the accuracy of azimuth estimation by setting section parameters according to the specifications of the bumper, and furthermore, can prevent an increase in the time required to collect data required to calculate the correction matrix, and also prevent an increase in the amount of calculation.

[0069] (2) The radar device 1 of the first embodiment further includes an external database 40 that stores in advance section parameter information corresponding to various bumper specifications. The memory unit 32 is configured to receive section parameter information corresponding to the specifications of a predetermined bumper provided on a vehicle on which the radar device 1 is mounted from the external database 40. According to this, after the radar device 1 is mounted on a vehicle, there may be cases where it is not possible to acquire section parameter information corresponding to a predetermined bumper equipped on the vehicle using corner reflectors, etc. Even in such cases, by storing section parameter information corresponding to various bumper specifications in the external database 40 in advance, the memory unit 32 can acquire section parameter information corresponding to the predetermined bumper specifications from the external database 40.

[0070] (3) In the first embodiment, the specifications of the bumper include the curvature of the bumper, the paint material of the bumper, and the number of times the bumper is painted. According to this, as a result of detailed investigation by the inventors, it was found that the section parameter information that enables highly accurate direction estimation varies depending on the bumper curvature, the bumper paint material, and the number of times the bumper has been painted. It was also found that the radar-bumper distance does not affect the section parameter information. Therefore, by including the bumper curvature, the bumper paint material, and the number of times the bumper has been painted as bumper specifications, it is possible to set section parameter information that enables highly accurate direction estimation.

[0071] (4) The signal processing method of the radar device 1 of the first embodiment includes the following processes: storing a correction matrix and section parameter information set for each correction section according to the specifications of the bumper in the memory unit 32 (S15); calculating a corrected array mode vector by performing array correction on the array mode vector using the correction matrix (S247); estimating the target direction using radio waves transmitted and received by the array antenna and the corrected array mode vector (S230); calculating the target direction from the relative speed between the vehicle and the target while the vehicle is moving (S242); storing peak data obtained by frequency analysis of radio waves received when calculating the target direction in a bin of the correction section corresponding to the target direction calculated from the relative speed (S244); calculating a correction matrix for a correction section among multiple correction sections in which peak data is stored in at least all of its own bins, and updating the correction matrix for that correction section (S246). According to this method, even if the radar-bumper distance changes while the vehicle is in use, the correction matrix of the correction section is updated accordingly. Therefore, the corrected array mode vector is calculated using the updated correction matrix. Therefore, this signal processing method can maintain or achieve high-accuracy target azimuth estimation.

[0072] (5) In the signal processing method of the radar device 1 of the first embodiment, storing the correction matrix set for each correction section according to the bumper specifications and the section parameter information in the memory unit 32 includes the following processes: 1. Storing the section parameter information corresponding to the specifications of various bumpers in the external database 40 in advance (S15); 2. Inputting the section parameter information corresponding to the specifications of a predetermined bumper provided on the vehicle on which the radar device 1 is mounted from the external database 40 to the memory unit 32. According to this, when the specifications of the bumper of the vehicle on which the radar device 1 is mounted are determined, the memory unit 32 can acquire, from the external database 40, the section parameter information corresponding to the specifications of the bumper.

[0073] (Second embodiment) Next, a second embodiment will be described with reference to Fig. 23. In the second embodiment, differences from the first embodiment will be described.

[0074] As shown in FIG. 23, in the second embodiment, the correction matrix Q k The width of the interval for collecting data required to calculate the correction matrix Q (hereinafter referred to as "calculation data") is set to be larger than the width of the correction interval. Specifically, the width of the interval for collecting calculation data is set to include the correction interval as well as the bins adjacent to both ends of the correction interval. k The amount of calculation data required varies depending on the number of antennas and the method for calculating the correction matrix. Therefore, it is possible to make the width of the section in which calculation data is collected wider than the width of the correction section.

[0075] In this case, in the second embodiment, in S245 described in the flowchart of FIG. 18, the control unit 30 performs processing to determine whether there is any free FFT peak data in the multiple bins in the corresponding section and in the bins adjacent to both ends of the corresponding section.

[0076] In the second embodiment described above, the amount of calculation data for calculating the correction matrix Q can be increased compared to the first embodiment, and therefore it is expected that the estimation error of the direction can be further reduced.

[0077] (Modification of the second embodiment) In the second embodiment, the width of the section for collecting calculation data is set to include the correction section and one bin adjacent to both ends of the correction section, but this is not limiting, and the width may be set to include multiple bins adjacent to both ends of the correction section in addition to the correction section.Furthermore, the width of the section for collecting calculation data may be set to include one or more bins adjacent to either end of the correction section in addition to the correction section.

[0078] (Third embodiment) Next, a third embodiment will be described with reference to Fig. 24. In the third embodiment as well, differences from the first embodiment and the like will be described.

[0079] As shown in Fig. 24, in the third embodiment, the angle step of the section parameter is set smaller than those described in the first and second embodiments. As a result, the correction matrix Q k It is possible to increase the number of calculation data required to calculate Furthermore, by changing the angle step for collecting calculation data for each correction interval, it is possible to increase the number of calculation data for that correction interval.

[0080] (Other embodiments) Representative embodiments of the present disclosure have been described above, but the present disclosure is not limited to the above-described embodiments and can be modified in various ways, for example, as follows.

[0081] In the above embodiment, the radar device 1 is described as being mounted on a vehicle as an example of a moving body, but the radar device 1 is not limited to this and may be mounted on a moving body other than a vehicle.

[0082] In the above embodiment, the radar device 1 is described as being mounted between the bumper as a protective member of the vehicle and the vehicle body, but this is not limited thereto, and the radar device 1 may also be mounted between a protective member such as a grille, emblem, etc. and the vehicle body.

[0083] The present disclosure is not limited to the above-described embodiments and can be modified as appropriate within the scope of the claims. Furthermore, the above-described embodiments and portions thereof are not unrelated to each other and can be combined as appropriate unless the combination is clearly impossible. It goes without saying that, in each of the above embodiments, the elements constituting the embodiments are not necessarily essential unless specifically stated as essential or clearly considered essential in principle. Furthermore, in each of the above embodiments, when numerical values ​​such as the number, values, amounts, and ranges of components of the embodiments are mentioned, they are not limited to the specific numbers unless specifically stated as essential or clearly limited to a specific number in principle. Furthermore, in each of the above embodiments, when the shape, positional relationship, etc. of components are mentioned, they are not limited to the shape, positional relationship, etc., unless specifically stated or limited to a specific shape, positional relationship, etc. in principle.

[0084] The controller and method of the present disclosure may be implemented in a special-purpose computer by configuring a processor and memory programmed to perform one or more functions embodied in a computer program. The controller and method of the present disclosure may be implemented in a special-purpose computer by configuring a processor with one or more dedicated hardware logic circuits. The controller and method of the present disclosure may be implemented in one or more special-purpose computers configured with a processor and memory programmed to perform one or more functions in combination with a processor configured with one or more hardware logic circuits. The computer program may also be stored in a computer-readable non-transitory tangible storage medium as instructions executed by a computer. [Explanation of symbols]

[0085] 1. Radar equipment 2. Vehicles (moving objects) 3 Bumper (protective material) 32 Memory section 35 Direction estimation processing unit 36 Direction error estimation processing unit 37 Correction matrix calculation section 38 Array Correction Unit

Claims

1. A radar device mounted on a moving body (2), transmitting and receiving radio waves from an array antenna through a protective member (3) to estimate the direction of a target present outside the moving body, a memory unit (32) for storing a correction matrix set for each correction interval that divides the range of the viewing angle of the array antenna in accordance with the specifications of the protective member in order to perform array correction for correcting an array mode vector that indicates the phase of radio waves transmitted and received by the array antenna, and section parameter information including an angle step that divides the correction interval; an array correction unit (38) that calculates a corrected array mode vector by performing the array correction on the array mode vector using the correction matrix; an azimuth estimation processing unit (35) for estimating the azimuth of the target using radio waves transmitted and received by the array antenna and the corrected array mode vector; an azimuth error estimation processing unit (36) that calculates the azimuth of the target from the relative speed between the moving body and the target while the moving body is moving, calculates an error from the azimuth of the target estimated by the azimuth estimation processing unit, and stores peak data obtained by frequency analysis of radio waves received when calculating the azimuth of the target for the bin corresponding to the azimuth of the target calculated from the relative speed among a plurality of bins obtained by dividing the correction section by the angle step; a correction matrix calculation unit (37) that calculates the correction matrix for a correction interval in which the peak data is stored in at least all of the bins that the correction unit has, among the plurality of correction intervals, and updates the correction matrix for the correction interval.

2. An external database (40) is further provided which stores in advance the section parameter information corresponding to the specifications of various types of the protective members, 2. The radar device according to claim 1, wherein the memory unit is configured to input the section parameter information corresponding to the specifications of the predetermined protective member provided in the mobile body on which the radar device is mounted from the external database.

3. the protective member is a bumper, 3. The radar device according to claim 1, wherein the specifications of the protection member include a curvature of the bumper, a paint material for the bumper, and the number of times the bumper has been painted.

4. A signal processing method for a radar device mounted on a moving body (2), transmitting and receiving radio waves from an array antenna through a protective member (3) to estimate the direction of a target present outside the moving body, comprising: storing, in a memory unit (S15), a correction matrix set for each correction interval obtained by dividing the range of the viewing angle of the array antenna in accordance with the specifications of the protective member in order to perform array correction for correcting an array mode vector indicating the phase of radio waves transmitted and received by the array antenna, and interval parameter information including an angle step for dividing the correction interval; calculating a corrected array mode vector by performing the array correction using the correction matrix on the array mode vector (S247); Estimating the azimuth of the target using the radio waves transmitted and received by the array antenna and the corrected array mode vector (S230); Calculating the azimuth of the target from the relative speed between the moving body and the target while the moving body is moving (S242); storing peak data obtained by frequency analysis of the radio waves received when calculating the target's azimuth, in the bin corresponding to the target's azimuth calculated from the relative speed, among a plurality of bins obtained by dividing the correction section by the angle step (S244); calculating the correction matrix for a correction interval in which the peak data is stored in at least all of the bins of the correction interval among the plurality of correction intervals, and updating the correction matrix for the correction interval (S246).

5. Storing the correction matrix and the section parameter information set for each correction section in accordance with the specifications of the protective member in the memory unit The section parameter information corresponding to the specifications of the various protective members is stored in advance in an external database (40) (S15), 5. The signal processing method according to claim 4, further comprising inputting the section parameter information corresponding to the specifications of the predetermined protective member provided in the mobile body on which the radar device is mounted from the external database to the memory unit.

6. the protective member is a bumper, 6. The signal processing method according to claim 4, wherein the specifications of the protective member include a curvature of the bumper, a paint material for the bumper, and the number of times the bumper has been painted.

Images