Azimuth estimation device and azimuth estimation method

Abstract

To provide an azimuth estimation device and an azimuth estimation method capable of improving an azimuth-dependent error.SOLUTION: A radar device 1 functioning as an azimuth estimation device includes: a storage section 32; and a processor 31 that corrects an ideal mode vector, and estimates an azimuth by using a correction result of the mode vector and a reception signal corresponding to reflection waves received by a reception section 20. The storage section 32 stores multiple calibration matrices Q having a matrix element value optimized for each calibration section divided into N (an integer of 2 or larger) pieces in a range of a visual field angle of the reception section 20. The processor 31 corrects the mode vector for each calibration section by using multiple calibration matrices Q stored in the storage section 32.SELECTED DRAWING: Figure 1

Description

[Technical field]

[0001] The present disclosure relates to a direction estimation device and a direction estimation method. [Background technology]

[0002] Conventionally, a method is known in which a predetermined number of signals are selected from received signals received from known arrival angles, and an error model matrix is ​​calculated to reduce the difference between the arrival angle estimated based on the selected signals and the known arrival angle (see, for example, Patent Document 1). [Prior art documents] [Patent documents]

[0003] [Patent Document 1] JP 2017-211241 A Summary of the Invention [Problem to be solved by the invention]

[0004] When an on-board radar device is installed inside the bumper of a vehicle, the transmitted waves from the radar device and the waves reflected by the target object are refracted and reflected by the bumper, causing errors in the amplitude and phase of the received signal at some arrival angles and mutual coupling errors between antennas. Such errors that depend on the arrival angle are hereinafter also called azimuth-dependent errors.

[0005] Patent Document 1 describes the use of an error model matrix to estimate the angle of arrival of a reflected wave as the azimuth of a target. However, since a single error model matrix is ​​applied to all azimuths, it is not possible to improve the azimuth-dependent error described above.

[0006] An object of the present disclosure is to provide a direction estimation device and a direction estimation method capable of improving a direction-dependent error. [Means for solving the problem]

[0007] The invention described in claim 1 is An azimuth estimation device that estimates an arrival angle of a wave reflected by a target to a radar device as the azimuth of the target, A memory unit (32); a correction unit (31a) that corrects a mode vector indicating a phase according to an arrival angle of a reflected wave; a direction estimating unit (31b) that estimates a direction by using a corrected mode vector that is a result of the correction of the mode vector by the correcting unit and a received signal corresponding to a reflected wave received by a receiving unit of the radar device, the storage unit stores a plurality of calibration matrices having matrix element values ​​optimized for each calibration interval divided into N, the N being an integer equal to or greater than 2, within the range of the viewing angle of the receiving unit; The correction unit corrects the mode vector for each calibration interval using a plurality of calibration matrices stored in the storage unit.

[0008] The invention described in claim 5 is A direction estimation method for estimating an angle of arrival of a wave reflected by a target to a radar device as a direction of the target, comprising: A correction step (S130, S240) of correcting an ideal mode vector indicating a phase according to an arrival angle of a reflected wave; a direction estimation step (S140, S240) of estimating a direction using a corrected mode vector that is a result of the correction of the mode vector in the correction step and a received signal corresponding to a reflected wave received by a receiving unit of the radar device, A calibration step (S270) of obtaining a plurality of calibration matrices having optimized matrix element values ​​for each calibration section divided into N sections, where N is an integer equal to or greater than 2, within the range of the viewing angle of the receiving unit; In the correction step, the mode vector is corrected for each calibration interval using the multiple calibration matrices obtained in the calibration step.

[0009] In this way, if the configuration is such that the angle of arrival of the reflected wave is estimated using a mode vector corrected by a calibration matrix optimized for each specified calibration interval, it is possible to reduce azimuth-dependent errors compared to when the same calibration matrix is ​​used for all directions.

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

[0011] [Figure 1] 1 is a schematic configuration diagram of a radar device according to a first embodiment. [Diagram 2] 5 is a flowchart showing a flow of an estimation process executed by a processing unit of the radar device according to the first embodiment. [Diagram 3] FIG. 2 is an explanatory diagram for explaining a receiving section of the radar device according to the first embodiment. [Figure 4] FIG. 1 is an explanatory diagram for explaining an overview of a direction estimation process. [Diagram 5] FIG. 2 is an explanatory diagram for explaining a measurement system for determining a calibration matrix. [Figure 6] FIG. 11 is an explanatory diagram for explaining a method for determining a calibration matrix. [Figure 7] 1 is an explanatory diagram for explaining an estimation error of a direction when a direction is estimated by applying a single calibration matrix to all directions. FIG. [Figure 8] FIG. 11 is an explanatory diagram for explaining a calibration section divided into a range of a viewing angle. [Figure 9] 3 is an explanatory diagram for explaining an estimation process executed by a control unit of the radar device of the first embodiment. FIG. [Figure 10] FIG. 4 is an explanatory diagram for explaining the relationship between a calibration interval and an interval width in the first embodiment. [Figure 11] 10 is an explanatory diagram for explaining an example of an estimation error of a direction when a direction is estimated by applying a calibration matrix set for each calibration interval. FIG. [Figure 12] 11 is an explanatory diagram for explaining another example of an estimation error of a direction when a direction is estimated by applying a calibration matrix set for each calibration interval. FIG. [Figure 13] 4 is a flowchart showing the flow of a calculation process executed by the calibration device. [Figure 14]13 shows an example of an estimation error when an orientation estimation process is performed using the first calibration matrix obtained by the first method. [Figure 15] 13 shows an example of an estimation error when an orientation estimation process is performed using the second calibration matrix obtained by the second method. [Figure 16] 4 shows the estimation error when the direction estimation process is performed using the first calibration matrix and the second calibration matrix. [Figure 17] 10 is an explanatory diagram for explaining an estimation error of a direction when a direction estimation process is performed using an optimal calibration matrix. FIG. [Figure 18] FIG. 11 is an explanatory diagram for explaining the relationship between a calibration interval and an interval width in the second embodiment. [Figure 19] FIG. 13 is an explanatory diagram for explaining the relationship between a calibration interval and an interval width according to a modified example of the second embodiment. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

[0012] Hereinafter, the embodiments of the present disclosure will be described with reference to the drawings. In the following embodiments, the same reference numerals are used for parts that are the same as or equivalent to those described in the preceding embodiments, and the description thereof may be omitted. In addition, in the embodiments, when only a part of the components is described, the components described in the preceding embodiments can be applied to the other parts of the components. In the following embodiments, each embodiment can be partially combined with each other, even if not specifically stated, as long as the combination is not particularly hindered.

[0013] (First embodiment) The present embodiment will be described with reference to Fig. 1 to Fig. 17. In the present embodiment, an example will be described in which the direction estimation device and direction estimation method of the present disclosure are applied to a radar device 1 that is mounted on a vehicle and detects various targets existing around the vehicle.

[0014] The radar device 1 is disposed, for example, inside the front part of the vehicle, and emits radio waves toward the front of the vehicle and receives radio waves reflected by a target in front of the vehicle to determine the distance to the target, the relative speed with respect to the vehicle, the direction with respect to the vehicle, etc.

[0015] The radar device 1 employs the FMCW method as a signal modulation method. The radar device 1 operates at a radio frequency in 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 1 is not limited to a frequency corresponding to millimeter waves, and may be a frequency other than millimeter waves.

[0016] As shown in Fig. 1, the radar device 1 includes a transmitting unit 10, a receiving unit 20, and a control unit 30. The transmitting unit 10 has an oscillator 11 and a modulator 12, and transmits a transmission wave from a transmitting antenna 13. The receiving unit 20 receives a reflected wave of the transmission wave reflected by a target using a plurality of receiving antennas 21. The receiving unit 20 processes the signal received by the receiving 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.

[0017] The control unit 30 constitutes an electronic control unit in the radar device 1. The radar device 1 is a microcomputer equipped with a processor 31 and a storage unit 32. The storage unit 32 may be, for example, a ROM, a RAM, or an EEPROM. Various functions of the microcomputer are realized by executing a computer program stored in the storage unit 32. In this example, the storage unit 32 corresponds to a non-transient tangible recording medium storing the computer program. Some or all of the functions executed by the processor 31 may be realized by a hardware circuit.

[0018] In the control unit 30 of this embodiment, the processor 31 corrects the mode vector indicating a phase according to the arrival angle of the reflected wave, and estimates the direction using the corrected mode vector, which is the result of the correction of the mode vector, and the received signal X corresponding to the reflected wave. In this embodiment, the configuration in the processor 31 that corrects the mode vector constitutes the correction unit 31a, and the configuration in the processor 31 that estimates the direction using the corrected mode vector and the received signal X constitutes the direction estimation unit 31b.

[0019] 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 a plurality of transmitting antennas 13. SIMO is an abbreviation for Single Input Multiple Output. MIMO is an abbreviation for Multiple Input Single Output. MISO is an abbreviation for Single Input Multiple 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.

[0020] The radar device 1 executes a target detection process in the control unit 30 to estimate the distance to the target, the speed of the target, and the direction of the target based on the signal reception result in the receiving unit 20. The target detection process executed by the control unit 30 will be described below with reference to Fig. 2. The process shown in Fig. 2 is executed by the control unit 30 periodically or irregularly when a chirp signal is transmitted from the transmitting antenna 13 at a predetermined transmission period.

[0021] As shown in FIG. 2, when the reflected waves from the target are received by the multiple receiving antennas 21, the control unit 30 acquires, in step S100, a received signal X from the receiving unit 20 when the reflected waves from the target are received by the multiple receiving antennas 21.

[0022] For example, as shown in FIG. 3, the receiver 20 of the radar device 1 is a linear array having L receiving antennas 21 (element positions d 1、 d 2 , d L ) in the received signal X at each receiving antenna 21. In this case, the received signal X at each receiving antenna 21 is not an ideal signal, and may contain 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 formula F10 in FIG. 4. In formula F10, the received signal X is represented as "X(t)".

[0023] Here, "T" in formulas F10 to F13 and F15 indicates transposition. "s(t)" in formula F12 is a signal vector indicating the complex amplitude of each of K arriving waves. "n(t)" in formulas F11 and F13 is a noise vector whose elements are noise components whose average is "0" and whose variance is thermal noise power at each antenna element. "a(θ K )” is the ideal mode vector corresponding to the Kth arriving wave, and “A” is “a(θ K The mode vector is a vector indicating the phase according to the angle of arrival of the reflected wave. Note that "λ" is the wavelength.

[0024] In addition, “Q” in the formulas F11 and F16 is a matrix indicating an error model in the case where various array errors are included. In this embodiment, the matrix indicating the error model is a calibration matrix Q that corrects the ideal mode vector.

[0025] When receiving signal X from receiver 20, controller 30 performs FFT processing in step S110. Controller 30 performs frequency analysis of receiving signal X using FFT to obtain the distance to the target. Controller 30 also performs FFT for each beat frequency component to obtain the Doppler frequency and obtains the relative velocity of the target based on the Doppler frequency. Note that FFT is an abbreviation for Fast Fourier Transform.

[0026] Next, the control unit 30 performs FFT peak processing in step S120. The control unit 30 detects, for example, a frequency bin at which a peak occurs in an RV map formed of the distance and speed obtained by the FFT processing.

[0027] Next, the control unit 30 performs an array correction process in step S130. The array correction process is a process for correcting the ideal mode vector to obtain a corrected mode vector taking into account various array errors and the like. In the array correction process, the ideal mode vector "a(θ)" is corrected using a calibration matrix Q to obtain a corrected mode vector "am(θ)". Details of the calibration matrix Q will be described later. In this embodiment, the array correction process corresponds to a "correction step" for correcting the ideal mode vector.

[0028] Next, the control unit 30 performs an azimuth estimation process in step S140. In this embodiment, the azimuth estimation process corresponds to a "azimuth estimation step" of estimating an azimuth using the corrected mode vector and the received signal X.

[0029] Specifically, in the direction estimation process, a correlation matrix R shown in formula F17 of FIG. 4 is calculated based on the received signal X. XX Then, the control unit 30 calculates the correlation matrix R XX From the eigenvalue decomposition of N Note that “T” in formula F17 is not the transpose but the number of snapshots.

[0030] Then, the control unit 30 calculates the noise subspace U Nand the correction mode vector “am(θk)” corrected by the calibration matrix Q are orthogonal (i.e., am(θk) ⊥ U N The direction is estimated by the MUSIC method shown in formula F19 in Fig. 4 using θ = θk (k = 1, 2, ... K). In this MUSIC method, when θ = θk, the denominator of formula F19 approaches zero and becomes a very large value. By detecting this peak value, the arrival angle of the reflected wave is estimated as the direction of the target. Note that 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, MODE, etc.

[0031] Here, a measurement system for determining the calibration matrix Q and the like will be described with reference to Fig. 5 to Fig. 9. In the measurement system for determining the calibration matrix Q, for example, as shown in Fig. 5, a calibration device 40, a rotator 50 for rotating the calibration device 40, and a corner reflector 60 serving as a target are used.

[0032] The calibration device 40 is a device having the same functions as the radar device 1. Note that the calibration device 40 may be a device other than the radar device 1 that has the same functions as the radar device 1, or the radar device 1 itself may be used.

[0033] The measurement system for determining the calibration matrix Q uses a rotator 50 to rotate the calibration device 40 at a predetermined angle, thereby setting the arrival angle θ of the reflected wave from the corner reflector 60, and the received signal X at that time can be measured by the calibration device 40. Note that Fig. 5 is a plan view, and the arrival angle θ is an angle in the horizontal direction. Also, the arrival angle θ in the direction perpendicular to the front of the calibration device 40 is set to zero.

[0034] When determining the calibration matrix Q using such a measurement system, the calibration device 40 is rotated in predetermined angular steps within the range of the viewing angle of the receiving unit 20, and the received signal X measured each time is stored as received data in the memory of the calibration device 40. Then, the calibration device 40 determines the calibration matrix Q based on the received data stored in the memory.

[0035] There are various methods for obtaining the calibration matrix Q, which can be roughly divided into "method A" and "method B" shown in Fig. 6. Method A uses a received signal Xmeans received from a known arrival angle, which is a known arrival angle, or a signal subspace U obtained by eigenvalue decomposition of the received signal Xmeans. S and the ideal mode vector determined by the known arrival angle. Method B uses the noise subspace U obtained by eigenvalue decomposition of the received signal Xmeans received from a known arrival angle, which is a known arrival angle. N and an ideal mode vector determined by a known arrival angle. The received signal Xmeans is the average of the actual received signal X over a predetermined number of snapshots.

[0036] Incidentally, for example, when the radar device 1 is installed inside the bumper of a vehicle, the transmitted wave from the radar device 1 and the reflected wave from the target may be refracted or reflected by the bumper. This effect may cause azimuth-dependent errors such as errors in the amplitude and phase of the received signal X at some angles of arrival and mutual coupling errors between antennas.

[0037] Fig. 7 is an explanatory diagram for explaining the estimation error of the direction when a single calibration matrix Q is applied to all directions. The calibration matrix Q in this example is calculated using the reception data (hereinafter also referred to as calibration data) when the calibration device 40 is rotated in the range of the viewing angle in angle steps of "1 [deg]" as shown in Fig. 7(a). In this example, the reception data when the calibration device 40 is rotated in the range of the viewing angle in angle steps of "0.5 [deg]" is used as the evaluation data, and the difference between the direction estimated based on the evaluation data and the known arrival angle is calculated as the estimation error of the direction.

[0038] When the azimuth of a target is estimated by applying a single calibration matrix Q to all azimuths, the estimation error of the azimuth tends to be large at some arrival angles, as shown in Fig. 7(b).In addition, the peak of the spectrum power tends to be small and the peak width tends to be large, as shown in Fig. 7(c).

[0039] In consideration of these, the radar device 1 of this embodiment divides the range of the viewing angle into a plurality of intervals, and corrects the mode vector by a calibration matrix Q set for each of the intervals, thereby improving the azimuth estimation error.

[0040] For example, as shown in FIG. 8, the radar device 1 uses a calibration matrix Q 1 ~Q N is stored in the memory unit 32 of the control unit 30. This calibration matrix Q 1 ~Q N is a matrix with the arrival angle θ as an argument, as shown in, for example, formula F26 in FIG.

[0041] The radar device 1 of this embodiment uses a plurality of calibration matrices Q 1 ~Q N 9, the mode vector is corrected for each calibration interval. After that, the radar device 1 estimates the azimuth of the target using a corrected mode vector, which is the result of the correction of the mode vector, and the reception signal X received by the receiver 20, as shown in, for example, formula F29 in Fig. 9. Note that formulas F20, F22 to F25, F27, and F28 shown in Fig. 9 are the same as formulas F10, F12 to F15, F17, and F18 shown in Fig. 4.

[0042] The calibration matrix Q of this embodiment 1 ~Q N For example, as shown in Fig. 10, is calculated using received data obtained at multiple points within the calibration interval. In Fig. 10, six calibration intervals are set within the range of the viewing angle, and six calibration matrices Q 1 ~Q6 This shows an example of what was sought.

[0043] In this embodiment, the width of the section in which calibration data is collected (hereinafter also referred to as section width) is the same as the width of the calibration section. Also, the angle step in which calibration data is collected is set to the same value in each calibration section. In the example shown in FIG. 10, the angle step is set to 2[deg].

[0044] Here, for example, under data conditions where the number of sections N in the calibration section is large or the angle step for collecting calibration data is small as shown in Fig. 11(a), the influence of the orientation-dependent error tends to be small as shown in Fig. 11(b) and Fig. 11(c). In such cases, the effect of reducing the orientation estimation error can be sufficiently obtained.

[0045] On the other hand, for example, as shown in FIG. 12(a), the angle step for collecting the calibration data is large and the calibration matrix Q 1 ~Q 6 When the number of received data is small, the influence of the direction-dependent error tends to be large, as shown in Fig. 12(b) and Fig. 12(c). In such a case, the effect of reducing the direction estimation error may be small.

[0046] In this way, the effect of reducing the estimation error of the orientation changes depending on the number of sections N of the calibration section and the angle step. The effect of reducing the estimation error of the orientation also changes depending on the method for calculating the section width and the calibration matrix Q.

[0047] In consideration of these, the calibration device 40 performs an optimization process to select optimal section parameters that affect the effect of reducing the azimuth estimation error, such as the number of sections N of the calibration section, the angle step, the section width, and the method of obtaining the calibration matrix Q, and optimize the calibration matrix Q. The optimization process performed by the calibration device 40 will be described below with reference to Fig. 13. The process shown in Fig. 13 is performed by the calibration device 40 when a chirp signal is transmitted from the transmitting antenna 13 at a predetermined transmission period.

[0048] 13, the calibration device 40 acquires a reception signal X from a known direction in step S200. Specifically, the calibration device 40 stores the reception signal X measured each time it is rotated by a predetermined angle step in a memory as reception data.

[0049] Next, in step S210, the calibration device 40 sets section parameters such as the number of sections N of the calibration section, the angle step, the section width, etc. For example, the calibration device 40 sets a predetermined initial value to the section parameters for the first time, and from the next time onwards, sets the section parameters to values ​​obtained by changing any one of the number of sections N, the angle step, and the section width.

[0050] Next, in step S220, the calibration device 40 executes a calculation process to obtain a first calibration matrix Qa by a first method for each calibration section. For example, the calibration device 40 determines one of the methods A listed in Fig. 6 as the first method, and obtains the first calibration matrix Qa based on that method. Note that in this embodiment, the process of step S220 corresponds to a "first calculation step of obtaining the first calibration matrix Qa."

[0051] Next, in step S230, the calibration device 40 executes a calculation process to obtain a second calibration matrix Qb for each calibration section by a second method different from the first method. For example, the calibration device 40 determines one of the methods B listed in Fig. 6 as the second method, and obtains the second calibration matrix Qb based on that method. Note that in this embodiment, the process of step S230 corresponds to a "second calculation step of obtaining the second calibration matrix Qb."

[0052] Next, in step S240, the calibration device 40 executes an orientation estimation process to estimate the orientation of the corner reflector 60. In this orientation estimation process, the orientation of the corner reflector 60 is estimated by, for example, FFT processing, FFT peak processing, array correction processing, and orientation estimation processing shown in Fig. 2. Specifically, in the orientation estimation process, an orientation estimation result using a mode vector corrected by a first calibration matrix Qa is obtained as a first estimated orientation, and an orientation estimation result using a mode vector corrected by a second calibration matrix Qb is obtained as a second estimated orientation. In this embodiment, the process of step S240 corresponds to a "correction step" and an "orientation estimation step".

[0053] Next, in step S250, calibration device 40 determines whether the estimation error of the direction obtained in the direction estimation process is equal to or smaller than a predetermined threshold value. In this determination process, not only the estimation error of the direction based on the received signal X used in obtaining the first calibration matrix Qa and the second calibration matrix Qb, but also the estimation error of the direction based on the received signal X not used in obtaining each calibration matrix Q are compared with the threshold value.

[0054] Furthermore, the calibration device 40 compares the first and second estimated directions obtained in the direction estimation process with the known arrival angles to obtain the estimation errors of the first and second estimated directions. The calibration device 40 then determines whether or not one of the estimation errors of the first and second estimated directions is equal to or smaller than a threshold. The threshold is set to, for example, the upper limit of the allowable range of direction errors.

[0055] If the estimation error exceeds the threshold value, the calibration device 40 returns to step S210, finds the calibration matrix Q with section parameters different from those used up until now, and performs the direction estimation process using the calibration matrix Q.

[0056] On the other hand, if the estimated error is equal to or less than the threshold, the calibration device 40 outputs, in step S260, the section parameters for which the estimated error is equal to or less than the threshold as optimal section parameters. Then, in step S270, the calibration device 40 outputs, from the first calibration matrix Qa and the second calibration matrix Qb, the one that reduces the azimuth estimation error, as the calibration matrix Q having optimized matrix element values. Note that in this embodiment, the process of step S270 corresponds to a "calibration step" for determining a plurality of optimized calibration matrices Q.

[0057] Specifically, the calibration device 40 sets the first calibration matrix Qa as the calibration matrix Q having optimized matrix element values ​​for a section where the estimation error when the first calibration matrix Qa is used is equal to or smaller than the estimation error when the second calibration matrix Qb is used. Furthermore, the calibration device 40 sets the second calibration matrix Qb as the calibration matrix Q having optimized matrix element values ​​for a section where the estimation error when the first calibration matrix Qa is used is larger than the estimation error when the second calibration matrix Qb is used. The calibration matrix Q thus obtained is stored in the memory unit 32 of the radar device 1 and is used when estimating the azimuth of the target.

[0058] Here, Fig. 14 shows an example of an estimation error when the orientation estimation process is performed using the first calibration matrix Qa obtained by the first method with optimal section parameters. Fig. 15 shows an example of an estimation error when the orientation estimation process is performed using the second calibration matrix Qb obtained by the second method with optimal section parameters. Fig. 16 shows an estimation error when the orientation estimation process is performed using each calibration matrix Q with optimal section parameters. In Fig. 16, the estimation error when the orientation estimation process is performed using the first calibration matrix Qa is indicated by a solid line, and the estimation error when the orientation estimation process is performed using the second calibration matrix Qb is indicated by a dashed dotted line.

[0059] 14 and 15, the estimation error when the first calibration matrix Qa is used is larger than the estimation error when the second calibration matrix Qb is used in a portion of the viewing angle range (a range of 20 to 40 degrees in this example) as shown in Fig. 16. In this case, the second calibration matrix Qb is set as the calibration matrix Q for a portion of the viewing angle range (a range of 20 to 40 degrees in this example), and the first calibration matrix Qa is set as the calibration matrix Q for the rest of the viewing angle. This makes it possible to reduce the estimation error over the entire viewing angle range, as shown in Fig. 17, for example.

[0060] The radar device 1 described above functions as an azimuth estimation device that estimates the azimuth of a target. The method of estimating the azimuth of a target using the calibration device 40 and the like corresponds to the azimuth estimation method disclosed herein.

[0061] The control unit 30 of the radar device 1 of this embodiment includes a storage unit 32 and a processor 31 that corrects a mode vector indicating a phase according to the arrival angle of a reflected wave and estimates an orientation using the correction result of the mode vector and a received signal X received by the receiving unit 20. The storage unit 32 stores a plurality of calibration matrices Q having matrix element values ​​optimized for each calibration interval divided into N, where N is an integer equal to or greater than 2, within the range of the field of view angle of the receiving unit 20. The processor 31 corrects the mode vector for each calibration interval using the plurality of calibration matrices Q stored in the storage unit 32.

[0062] In this way, if the angle of arrival of the reflected wave is estimated using a mode vector corrected by a calibration matrix Q optimized for each specified calibration interval, it is possible to reduce azimuth-dependent errors compared to when the same calibration matrix Q is used for all directions.

[0063] Moreover, according to this embodiment, the following effects can be obtained.

[0064] (1) In the storage unit 32, of the first calibration matrix Qa obtained by the first method and the second calibration matrix Qb obtained by the second method different from the first method, the one that reduces the azimuth estimation error is stored as the calibration matrix Q having optimized matrix element values. In this way, if the first calibration matrix Qa and the second calibration matrix Qb are obtained by different methods and the one that reduces the azimuth estimation error is adopted as the calibration matrix Q, it is possible to realize correction of the mode vector using the calibration matrix Q that is optimized for each calibration interval.

[0065] (2) The first method is to use a received signal X received from a known arrival angle, which is a known arrival angle, or a signal subspace U obtained by eigenvalue decomposition of the received signal X. S The second method uses the noise subspace U obtained by eigenvalue decomposition of the received signal X to obtain the first calibration matrix Qa. N and an ideal mode vector determined by a known angle of arrival. This makes it possible to realize correction of the mode vector using a calibration matrix Q optimized for each calibration interval.

[0066] (3) The storage unit 32 stores a calibration matrix Q having optimized matrix element values ​​that reduce the estimation error when estimating the direction based on the received signal X that is not used when determining the first calibration matrix Qa and the second calibration matrix Qb. In this way, by optimizing the matrix element values ​​of the calibration matrix Q so as to reduce the estimation error when estimating the direction based on the received signal X that is not used when determining each calibration matrix Q, it is possible to appropriately suppress the direction-dependent error.

[0067] Second embodiment Next, a second embodiment will be described with reference to Fig. 18. In this embodiment, differences from the first embodiment will be mainly described.

[0068] 18, in the second embodiment, the width of the section for collecting calibration data is wider than the width of the calibration section. Specifically, the width of the section in this embodiment is set to include a plurality of points adjacent to both ends of the calibration section in addition to the calibration section.

[0069] In the second embodiment, the angle step for collecting the calibration data is set to the same value in each calibration interval, as in the first embodiment. In the example shown in Fig. 18, the angle step is set to 2 [deg].

[0070] The calibration matrix Q of this embodiment is determined using received signals X obtained at multiple points within the section width. This allows for more received data when calculating the calibration matrix Q than in the first embodiment, and is expected to further reduce the estimation error of the direction.

[0071] The rest of the configuration is the same as in the first embodiment. The direction estimation device and the direction estimation method of the present embodiment can obtain the same effects as in the first embodiment, which are achieved by a configuration common to the first embodiment or an equivalent configuration.

[0072] (Modification of the second embodiment) In the second embodiment, the angle step for collecting calibration data is set to 2 [deg] as an example, but the angle step is set to an angle calculated by the optimization process. In the second embodiment, the section width for collecting calibration data is expanded more than the width of the calibration section, and the amount of data received when calculating the calibration matrix Q is increased, so that the angle step may be set to 3 [deg], for example, as shown in FIG. 19. Note that the angle step does not have to be the same value in each calibration section, and may be changed depending on the position of the calibration section. This is also true for the first embodiment.

[0073] In the second embodiment, the calibration section and a plurality of points adjacent to both ends of the calibration section are set as the section width, but this is not limited to this. The section width may be set to include the calibration section and a plurality of points adjacent to one end of the calibration section. Note that the section width may be changed depending on the position of the calibration section.

[0074] (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 described below.

[0075] In the above embodiment, the calibration matrix Q that reduces the azimuth estimation error between the first calibration matrix Qa obtained by the first method and the second calibration matrix Qb obtained by the second method is stored in the storage unit 32 as the calibration matrix Q, but the present invention is not limited to this. For example, the radar device 1 may store in the storage unit 32 a calibration matrix Q obtained by only one of the first method and the second method.

[0076] As described in the above embodiment, it is desirable to determine the optimum calibration matrix Q as one that reduces the estimation error when estimating the direction based on the received signal X that was not used when determining the first calibration matrix Qa and the second calibration matrix Qb, but this is not limiting. The direction estimation device and direction estimation method may be configured to determine the optimum calibration matrix Q as one that reduces the estimation error when estimating the direction based on the received signal X that was used when determining the first calibration matrix Qa and the second calibration matrix Qb.

[0077] In the above embodiment, the direction estimation device and the direction estimation method of the present disclosure are applied to the radar device 1 mounted on a vehicle to detect various targets around the vehicle, but the present disclosure is not limited to this. The direction estimation device and the direction estimation method of the present disclosure can also be applied to a radar device 1 installed on equipment or objects other than a vehicle.

[0078] In the above-described embodiments, it goes without saying that the elements constituting the embodiments are not necessarily essential, except in cases where it is particularly expressly stated that they are essential or where they are clearly considered to be essential in principle.

[0079] In the above-described embodiments, when numerical values ​​such as the number, values, amounts, ranges, etc. of components of the embodiments are mentioned, they are not limited to the specific numbers unless it is specifically stated as required or unless it is clearly limited to a specific number in principle.

[0080] In the above-described embodiments, when referring to the shapes, positional relationships, etc. of components, etc., there is no limitation to those shapes, positional relationships, etc., unless specifically stated or in principle limited to a specific shape, positional relationship, etc.

[0081] The control unit and the method of the present disclosure may be implemented in a special-purpose computer provided by configuring a processor and a memory programmed to execute one or more functions embodied in a computer program. The control unit and the method of the present disclosure may be implemented in a special-purpose computer provided by configuring a processor with one or more dedicated hardware logic circuits. The control unit and the method of the present disclosure may be implemented in one or more special-purpose computers configured by a combination of a processor and a memory programmed to execute one or more functions and a processor configured with one or more hardware logic circuits. The computer program may also be stored in a computer-readable non-transitory tangible recording medium as instructions executed by a computer. [Explanation of symbols]

[0082] 1. Radar equipment 31a Correction section 31b Orientation estimation part 32 Storage section

Claims

1. A direction estimation device that estimates an angle of arrival of a wave reflected by a target to a radar device (1) as the direction of the target, comprising: A memory unit (32); A correction unit (31 a) that corrects a mode vector indicating a phase according to the arrival angle of the reflected wave; an azimuth estimation unit (31b) that estimates the azimuth by using a corrected mode vector that is a correction result of the mode vector by the correction unit and a reception signal corresponding to the reflected wave received by a reception unit of the radar device, the storage unit stores a plurality of calibration matrices having matrix element values ​​optimized for each calibration interval divided into N, where N is an integer equal to or greater than 2, within a range of a viewing angle of the receiving unit; The correction unit corrects the mode vector for each calibration interval using the multiple calibration matrices stored in the storage unit.

2. 2. The orientation estimation device according to claim 1, wherein the storage unit stores, as the calibration matrix having the optimized matrix element values, one of a first calibration matrix calculated by a first method and a second calibration matrix calculated by a second method different from the first method, which reduces an estimation error of the orientation.

3. The first method is a method of obtaining the first calibration matrix by using the received signal received from a known arrival angle, which is a known arrival angle, or a signal subspace obtained by eigenvalue decomposition of the received signal, and an ideal mode vector determined by the known arrival angle, 3. The direction estimation device according to claim 2, wherein the second method is a method for determining the second calibration matrix using a noise subspace obtained by eigenvalue decomposition of the received signal and the ideal mode vector determined by the known arrival angle.

4. 4. The direction estimation device according to claim 3, wherein the storage unit stores, as the calibration matrix having the optimized matrix element values, a matrix that reduces the estimation error when the direction is estimated based on the received signals that are not used when determining the first calibration matrix and the second calibration matrix.

5. A method for estimating an angle of arrival of a wave reflected by a target to a radar device (1) as the azimuth of the target, comprising: A correction step (S130, S240) of correcting a mode vector indicating a phase according to the arrival angle of the reflected wave; a direction estimating step (S140, S240) of estimating the direction by using a corrected mode vector that is a correction result of the mode vector in the correction step and a reception signal corresponding to the reflected wave received by a receiving unit of the radar device, and a calibration step (S270) of obtaining a plurality of calibration matrices having optimized matrix element values ​​for each calibration interval divided into N, where N is an integer equal to or greater than 2, within the range of the viewing angle of the receiving unit; In the correcting step, the mode vector is corrected for each calibration interval using the multiple calibration matrices obtained in the calibration step.

6. A first calculation step (S220) of obtaining a first calibration matrix using the received signal received from a known arrival angle, which is a known arrival angle, or a signal subspace obtained by eigenvalue decomposition of the received signal, and an ideal mode vector determined by the known arrival angle; A second calculation step (S230) of obtaining a second calibration matrix using a noise subspace obtained by eigenvalue decomposition of the received signal and the ideal mode vector determined by the known arrival angle, 6. The orientation estimation method according to claim 5, wherein in the calibration step, one of the first calibration matrix and the second calibration matrix that reduces an estimation error of the orientation is determined as the calibration matrix having the optimized matrix element values.

7. 7. The direction estimation method according to claim 6, wherein in the calibration step, a calibration matrix having the optimized matrix element values ​​is obtained that reduces the estimation error when the direction is estimated based on the received signals that are not used when determining the first calibration matrix and the second calibration matrix.

Images