Axis offset estimation device

Abstract

The invention relates to a shaft offset estimation device. The estimation device (6) extracts, from among a plurality of first observation points detected by the radar device (2), a first observation point that is reflected by a roadside object that is arranged in the direction in which the travel path extends at a position higher than the travel path on the side of the travel path on which the moving body travels, as a first roadside object observation point. The estimation device extracts a second roadside object observation point, which is a second observation point corresponding to the first roadside object observation point, from a plurality of second observation points detected by the radar device on the basis of a preset correlation condition. The estimation device calculates, on the basis of the distribution of the positions of the plurality of extracted second roadside object observation points, direction information (beta) indicating the direction in which the roadside object extends, and calculates, as an axis offset angle, the angle at which the central axis (CA) of the radar device is inclined with respect to the front-rear direction of the moving body.

Description

[0001] CROSS-REFERENCE TO RELATED APPLICATIONS

[0002] This international application claims the benefit of priority based on Japanese Patent Application No. 2023-024391 filed with the Japan Patent Office on February 20, 2023, the entire contents of which are incorporated herein by reference. Technical Field

[0003] The present disclosure relates to an axis offset estimation device for estimating an axis offset angle of a radar device. Background Art

[0004] Patent document 1 describes an axial offset estimation device that obtains multiple roadside object information from a radar device by sending radar waves toward roadside objects arranged on the side of the road along the extension direction of the road on which a mobile body is traveling, and estimates the axial offset angle of the radar device based on the multiple roadside object information.

[0005] Patent Document 1: Japanese Patent Application Laid-Open No. 2021-148561

[0006] As a result of detailed research by the inventors, they discovered that in order to improve the object detection accuracy of a radar device, it is necessary to further improve the estimation accuracy of the axis offset angle of the radar device. Summary of the Invention

[0007] The present disclosure improves the estimation accuracy of the shaft offset angle.

[0008] One aspect of the present disclosure is an axial offset estimation device for estimating an axial offset angle of a radar device mounted on a mobile object. The device includes a first roadside object extraction unit, a second roadside object extraction unit, and an axial offset angle calculation unit.

[0009] The radar device is configured to detect a first observation point by transmitting and receiving a first radar wave modulated by a first modulation method, and repeatedly output first observation point information including a first observation point distance and a first observation point azimuth, wherein the first observation point is a location where the first radar wave is reflected, the first observation point distance is the distance between the radar device and the first observation point, and the first observation point azimuth is the azimuth at which the first observation point is located.

[0010] The radar device is configured to detect a second observation point by transmitting and receiving a second radar wave modulated by a second modulation method different from the first modulation method, and repeatedly output second observation point information including a second observation point distance and a second observation point azimuth, wherein the second observation point is a location at which the second radar wave is reflected, the second observation point distance is the distance between the radar device and the second observation point, and the second observation point azimuth is the azimuth at which the second observation point is located.

[0011] The first roadside object extraction unit is configured to extract a first observation point reflected by a roadside object from multiple first observation points detected by the radar device as a first roadside object observation point. The above-mentioned roadside object is arranged on the side of the driving path on which the mobile body is traveling at a position higher than the driving path and along the direction in which the driving path extends.

[0012] The second roadside object extraction unit is configured to extract a second roadside object observation point from multiple second observation points detected by the radar device based on a pre-set association condition indicating that the position of the first roadside object observation point is close to the position of the second observation point. The second roadside object observation point is a second observation point corresponding to the first roadside object observation point.

[0013] The axis offset angle calculation unit is configured to calculate the angle at which the central axis representing the direction in which the first radar wave and the second radar wave are sent and received by the radar device is tilted relative to the front-rear direction of the mobile body by calculating the direction information representing the direction in which the roadside object extends based on the distribution of the positions of multiple second roadside object observation points extracted by the second roadside object extraction unit.

[0014] The axial deviation estimation device of the present invention constructed in this way applies a modulation method suitable for detecting roadside objects arranged in the direction extending along the driving road to the first modulation method, and applies a modulation method with higher detection accuracy of the position of the observation point to the second modulation method, so that the direction information representing the direction in which the roadside objects extend can be calculated with better accuracy, thereby improving the estimation accuracy of the axial deviation angle. BRIEF DESCRIPTION OF THE DRAWINGS

[0015] Figure 1 This is a block diagram showing the structure of the shaft deviation detection system.

[0016] Figure 2 This is a diagram showing the modulation of radar waves.

[0017] Figure 3 It is a graph showing the maximum detection distances of the first and second modulations.

[0018] Figure 4 It is a diagram showing the structures of the first and second array antennas.

[0019] Figure 5 This is a flowchart showing the axis offset adjustment process.

[0020] Figure 6 4 is a flowchart showing the roadside object candidate point extraction process.

[0021] Figure 7 This is a flowchart showing the roadside object point cloud extraction process.

[0022] Figure 8This figure explains the method of extracting the roadside object point cloud.

[0023] Figure 9 This is a flowchart showing the process of associating roadside object point clouds.

[0024] Figure 10 : is a flowchart showing the axis offset angle estimation process according to the first embodiment.

[0025] Figure 11 It is a graph showing the angle of deviation from the vertical axis and the approximate straight line.

[0026] Figure 12 This is a diagram comparing the deviation in the position of the observation point between the first modulation and the second modulation.

[0027] Figure 13 : is a flowchart showing the axis deviation angle estimation process according to the second embodiment.

[0028] Figure 14 This is a diagram for explaining a method of calculating an approximate straight line in the second embodiment.

[0029] Figure 15 : is a flowchart showing the axis offset angle estimation process according to the third embodiment. DETAILED DESCRIPTION

[0030] [First embodiment]

[0031] The following, with the attached Figure 1 The first embodiment of the present disclosure will be described below.

[0032] like Figure 1 As shown, the axial misalignment detection system 1 of the present embodiment includes a radar device 2 , a radar mounting angle adjustment device 3 , a camera 4 , an on-vehicle sensor group 5 , and a control device 6 .

[0033] The radar device 2 is a MIMO radar that uses multiple antennas to simultaneously transmit and receive radio waves. MIMO is an abbreviation for Multi Input Multi Output.

[0034] Radar device 2 is installed on the front left side of a vehicle (hereinafter referred to as the host vehicle) equipped with axle misalignment detection system 1. Radar device 2 may be installed on the front right side of the host vehicle, on the rear left side of the host vehicle, on the rear right side of the vehicle, in front of the host vehicle, or behind the host vehicle.

[0035] The radar device 2 is arranged so that its detection range includes a front direction along the traveling direction of the host vehicle and a lateral direction perpendicular to the traveling direction.

[0036] The radar mounting angle adjustment device 3 includes a motor and a gear mounted on the radar device 2. The radar mounting angle adjustment device 3 rotates the motor in response to a drive signal output from the control device 6, thereby transmitting the rotational force to the gear, thereby rotating the radar device 2 about an axis extending along the vehicle width direction.

[0037] The camera 4 is mounted on the front side of the vehicle and continuously captures images of the situation in front of the vehicle.

[0038] The onboard sensor group 5 is a set of multiple sensors mounted on the vehicle to detect the vehicle's status. The onboard sensor group 5 includes a vehicle speed sensor that detects the vehicle's speed, an acceleration sensor that detects the vehicle's acceleration, and a yaw rate sensor that detects the vehicle's yaw rate.

[0039] The control device 6 is an electronic control unit centered around a microcomputer comprising a CPU 11, ROM 12, and RAM 13. The various functions of the microcomputer are realized by the CPU 11 executing a program stored on a non-migrating physical recording medium. In this example, the ROM 12 corresponds to the non-migrating physical recording medium storing the program. Furthermore, the execution of the program executes the corresponding method. Alternatively, some or all of the functions performed by the CPU 11 may be implemented in hardware using one or more integrated circuits (ICs). The number of microcomputers comprising the control device 6 may be one or more.

[0040] like Figure 2 As shown, radar device 2 switches between a first modulation period for generating a first high-frequency signal and a second modulation period for generating a second high-frequency signal within a predetermined modulation period Tm. The first high-frequency signal is composed of multiple first chirp signals with linearly increasing frequencies, and the second high-frequency signal is composed of multiple second chirp signals with linearly increasing frequencies. Furthermore, radar device 2 transmits the generated first and second high-frequency signals as first and second radar waves, respectively, and receives the reflected first and second radar waves. In this embodiment, the frequency width of the first chirp signal is smaller than the frequency width of the second chirp signal.

[0041] like Figure 3 As shown, radar apparatus 2 is configured such that the maximum detection distance (hereinafter, first maximum detection distance) of the modulation during the first modulation period (hereinafter, first modulation) is longer than the maximum detection distance (hereinafter, second maximum detection distance) of the modulation during the second modulation period (hereinafter, second modulation). Furthermore, radar apparatus 2 is configured such that the ranging accuracy of the second modulation is higher than the ranging accuracy of the first modulation.

[0042] like Figure 4 As shown, the radar device 2 includes a first array antenna 21 and a second array antenna 22 .

[0043] The first array antenna 21 is configured such that twelve antennas having the same characteristics are arranged at equal intervals in the horizontal direction (ie, vehicle width direction) and two antennas are arranged at equal intervals in the vertical direction (ie, vehicle height direction).

[0044] The second array antenna 22 is configured such that eight antennas having the same characteristics are arranged at equal intervals horizontally (i.e., vehicle widthwise) and four antennas are arranged at equal intervals vertically (i.e., vehicle heightwise). Radar device 2 uses first array antenna 21 to transmit and receive first radar waves during the first modulation period and second array antenna 22 to transmit and receive second radar waves during the second modulation period.

[0045] Therefore, the vertical angle measurement accuracy of the second modulation (hereinafter referred to as the second vertical angle measurement accuracy) is higher than the vertical angle measurement accuracy of the first modulation (hereinafter referred to as the first vertical angle measurement accuracy). In addition, the horizontal angle measurement accuracy of the first modulation is higher than the horizontal angle measurement accuracy of the second modulation.

[0046] The radar device 2 calculates the received power W of the received radar wave, the distance R to the point where the radar wave is reflected (hereinafter referred to as the observation point), the relative speed V to the observation point, the horizontal azimuth angle θ of the observation point, and the vertical azimuth angle θ of the observation point during each first modulation period and each second modulation period. Then, the radar device 2 will display the detected received power W, distance R, relative speed V, horizontal azimuth angle θ and vertical azimuth angle The observation point information is output to the control device 6. Hereinafter, the location where the first radar wave is reflected is referred to as the first modulation observation point, and the location where the second radar wave is reflected is referred to as the second modulation observation point. The observation point information obtained by the first modulation is referred to as the first modulation observation point information, and the observation point information obtained by the second modulation is referred to as the second modulation observation point information.

[0047] Next, a description will be given of the procedure of the axial offset adjustment process executed by the control device 6. The axial offset adjustment process is a process that is repeatedly executed every time the modulation period Tm elapses during the operation of the control device 6.

[0048] When performing axis offset adjustment processing, such as Figure 5 As shown, in S10 , the CPU 11 of the control device 6 executes a roadside object candidate point extraction process described later. The roadside object candidate point extraction process is a process of extracting roadside object candidate points that are candidates for roadside objects from a large amount of first modulated observation point information acquired from the radar device 2 .

[0049] In S20 , the CPU 11 executes a roadside object point cloud extraction process described later. The roadside object point cloud extraction process is a process of extracting a point cloud constituting a roadside object (hereinafter referred to as a roadside object point cloud) from the plurality of roadside object candidate points extracted in S10 .

[0050] In S30 , the CPU 11 executes a roadside object point cloud association process described later. The roadside object point cloud association process is a process of associating the roadside object point cloud extracted in S20 with the second modulated observation point.

[0051] In S40 , the CPU 11 executes an axis offset angle estimation process described later. The axis offset angle estimation process is a process for calculating the vertical axis offset angle θm of the radar device 2 based on the second modulation observation point associated in S30 .

[0052] In S50, the CPU 11 determines whether axis offset adjustment can be performed by the radar mounting angle adjustment device 3. Specifically, the CPU 11 determines whether the vertical axis offset angle θm calculated in S40 is less than or equal to a preset adjustable angle. If the vertical axis offset angle θm is less than or equal to the adjustable angle, the CPU 11 determines that axis offset adjustment can be performed.

[0053] Here, when axis offset adjustment can be performed, in S60, CPU11 uses the radar mounting angle adjustment device 3 to rotate the radar device 2 around the axis along the vehicle width direction of the vehicle by an axis offset angle θm to adjust the radar mounting angle so that the center axis CA of the radar device 2 is consistent with the front and rear direction of the vehicle, thereby ending the axis offset adjustment processing.

[0054] On the other hand, if the axis offset adjustment cannot be performed, in S70 , the CPU 11 outputs diagnostic information indicating that the central axis CA of the radar device 2 has shifted to the outside of the control device 6 , and ends the axis offset adjustment process.

[0055] Next, the steps of the roadside object candidate point extraction process executed in S10 will be described.

[0056] When performing the roadside object candidate point extraction process, as shown in Figure 6 As shown, in S110 , the CPU 11 selects one piece of first modulated observation point information not selected in the current roadside object candidate point extraction process from the first modulated observation point information newly acquired during the period from the previous roadside object candidate point extraction process to the current roadside object candidate point extraction process.

[0057] In S120, the CPU 11 determines whether a predetermined distance extraction condition is met for the first modulation observation point information selected in S110 (hereinafter, "selected observation point information"). In this embodiment, the distance extraction condition is that the distance R of the selected observation point information is greater than or equal to a predetermined first distance threshold and less than or equal to a predetermined second distance threshold. In this embodiment, the first distance threshold is, for example, 2 meters, and the second distance threshold is, for example, 100 meters.

[0058] If the distance extraction condition is not met, the CPU 11 proceeds to S190. On the other hand, if the distance extraction condition is met, the CPU 11 determines in S130 whether a preset horizontal azimuth extraction condition is met for the selected observation point information. In this embodiment, the horizontal azimuth extraction condition is that the horizontal azimuth angle θ of the selected observation point information is greater than or equal to a preset first horizontal azimuth angle threshold and less than or equal to a preset second horizontal azimuth angle threshold.

[0059] If the horizontal azimuth extraction condition is not met, the CPU 11 proceeds to S190. On the other hand, if the horizontal azimuth extraction condition is met, the CPU 11 determines in S140 whether a predetermined power extraction condition is met for the selected observation point information. In this embodiment, the power extraction condition is that the received power W of the selected observation point information is greater than or equal to a predetermined first power threshold and less than or equal to a predetermined second power threshold.

[0060] If the power extraction condition is not met, the CPU 11 proceeds to S190. On the other hand, if the power extraction condition is met, the CPU 11 determines in S150 whether a predetermined relative speed extraction condition is met for the selected observation point information. In this embodiment, the relative speed extraction condition is that the difference between the absolute value of the relative speed V in the selected observation point information and the absolute value of the traveling speed detected by the vehicle speed sensor included in the onboard sensor group 5 is less than a predetermined relative speed threshold. Furthermore, the relative speed threshold is set to indicate that the difference between the absolute value of the relative speed V and the absolute value of the traveling speed detected by the vehicle speed sensor is small.

[0061] Here, if the relative speed extraction condition is not met, CPU 11 moves to S190. On the other hand, if the relative speed extraction condition is met, in S160, CPU 11 determines whether the pre-set vehicle state extraction condition is met based on the selected observation point information. The vehicle state extraction condition of this embodiment is that both the pre-set acceleration extraction condition and the pre-set yaw rate extraction condition are met. The acceleration extraction condition of this embodiment is that the acceleration detected by the acceleration sensor included in the vehicle-mounted sensor group 5 is lower than the pre-set acceleration threshold. The yaw rate extraction condition of this embodiment is that the yaw rate detected by the yaw rate sensor included in the vehicle-mounted sensor group 5 is lower than the pre-set yaw rate threshold. In other words, the vehicle state extraction condition is that the vehicle is traveling in a straight line at a constant speed.

[0062] If the vehicle state extraction condition is not met, the CPU 11 proceeds to S190. On the other hand, if the vehicle state extraction condition is met, the CPU 11 determines in S170 whether the pre-set camera extraction condition is met for the selected observation point information. The camera extraction condition in this embodiment requires the presence of a roadside object at a location corresponding to the selected observation point information within the image captured by camera 4. The CPU 11 performs known image processing on the image captured by camera 4 to identify the roadside object within the image captured by camera 4.

[0063] Here, if the camera extraction condition is not satisfied, the CPU 11 proceeds to S190 . On the other hand, if the camera extraction condition is satisfied, the CPU 11 classifies the selected observation point information as a “roadside object candidate point” in S180 and proceeds to S200 .

[0064] When the process moves to S190 , the CPU 11 classifies the selected observation point information as “non-roadside object” and moves to S200 .

[0065] When the process proceeds to S200, the CPU 11 determines whether all of the newly acquired first modulation observation point information has been selected in S110. If not all of the first modulation observation point information has been selected, the CPU 11 proceeds to S110. On the other hand, if all of the first modulation observation point information has been selected, the CPU 11 terminates the roadside object candidate extraction process.

[0066] Next, the steps of the roadside object point cloud extraction process executed in S20 will be described.

[0067] When performing roadside object point cloud extraction processing, such as Figure 7As shown, in S310 , the CPU 11 performs candidate point clustering processing to group the plurality of roadside object candidate points. Specifically, the CPU 11 groups the plurality of roadside object candidate points into a plurality (e.g., 6) clusters based on their positions using, for example, the well-known k-means method.

[0068] In S320, the CPU 11 determines whether a pre-set longitudinal distance extraction condition is met for each of the multiple clusters generated in S310, and removes clusters for which the longitudinal distance extraction condition is not met. The longitudinal distance extraction condition in this embodiment is that the cluster has a length greater than a pre-set longitudinal distance threshold along the direction of travel of the vehicle. In this embodiment, the longitudinal distance threshold is, for example, 40 meters. Specifically, among the multiple roadside object candidate points constituting the cluster, if the difference between the distance R of the observation point information of the roadside object candidate point farthest from the vehicle along the direction of travel and the distance R of the observation point information of the roadside object candidate point closest to the vehicle along the direction of travel is greater than the longitudinal distance threshold, the CPU 11 determines that the longitudinal distance extraction condition is met.

[0069] In S330, CPU 11 determines whether a pre-set lateral distance extraction condition is met for clusters not removed in S320, and removes clusters for which the lateral distance extraction condition is not met. The lateral distance extraction condition in this embodiment is that the cluster's length along the vehicle width is less than a pre-set lateral distance threshold. In this embodiment, the lateral distance threshold is, for example, 1 meter. Specifically, if the difference between the distance R between the observation point information of the roadside object candidate point farthest from the vehicle along the vehicle width and the observation point information of the roadside object candidate point closest to the vehicle along the vehicle width is less than the lateral distance threshold, CPU 11 determines that the lateral distance extraction condition is met.

[0070] At S340 , the CPU 11 determines whether the pre-set lateral position extraction conditions are met for the clusters not removed in S320 and S330 . The CPU 11 sets the clusters for which the lateral position extraction conditions are met as "roadside object point clouds," terminating the roadside object point cloud extraction process. In this embodiment, the lateral position extraction condition is that the clusters are located on the innermost side of the left side of the vehicle.

[0071] like Figure 8 As shown, it is assumed that the host vehicle VH0 is traveling in a single-lane, two-lane passing lane LN1. In front of the host vehicle VH0, on the left side, three vehicles VH1, VH2, and VH3 are traveling in a lane LN2 in a straight line along the direction of travel. In addition, it is assumed that a guardrail GR is installed on the left side of the lane LN2.

[0072] It is assumed that the radar device 2 mounted on the vehicle VH0 transmits and receives the first modulated radar wave (i.e., the first radar wave mentioned above), so that the observation points OP1, OP2, OP3, OP4, OP5, OP6, OP7, OP8, OP9, OP10, OP11, OP12, OP13, OP14, and OP15 are detected in the order from near to far from the radar device 2.

[0073] The observation point OP1 is a point where the first radar wave is reflected on the rear right side of the vehicle VH1 , and the observation point OP2 is a point where the first radar wave is reflected on the front right side of the vehicle VH1 .

[0074] The observation point OP3 is a point where the first radar wave is reflected on the rear right side of the vehicle VH2 . The observation point OP4 is a point where the first radar wave is reflected on the front right side of the vehicle VH2 .

[0075] The observation point OP5 is a point where the first radar wave is reflected on the rear right side of the vehicle VH3 . The observation point OP6 is a point where the first radar wave is reflected on the front right side of the vehicle VH3 .

[0076] The observation points OP7 to OP15 are locations where the first radar wave is reflected by the guardrail GR.

[0077] Through the candidate point clustering process described above, a cluster CL1 including observation points OP1 to OP6 and a cluster CL2 including observation points OP7 to OP15 are generated.

[0078] Cluster CL1 is short in the direction of travel of the vehicle, so the longitudinal distance extraction condition is not met and it is removed in the process of S320 in the roadside object point cloud extraction process. Cluster CL2 is long in the direction of travel of the vehicle, so the longitudinal distance extraction condition is met and it is not removed in the process of S320 in the roadside object point cloud extraction process. Then, cluster CL2 is set as the "roadside object point cloud". Figure 8 In FIG, a solid-line rectangle indicates that a cluster is set as a “roadside object point cloud”, and a dotted-line rectangle indicates that a cluster is not set as a “roadside object point cloud”.

[0079] Furthermore, it is assumed that the radar device 2 mounted on the vehicle VH0 transmits and receives second-modulated radar waves (i.e., the aforementioned second radar waves), thereby detecting observation points OP21, OP22, OP23, OP24, OP25, OP26, OP27, OP28, OP29, and OP30 in descending order from the radar device 2. The maximum detection range of the second modulation is shorter than the maximum detection range of the first modulation, and the radar device 2 cannot detect observation points farther than the observation point OP30 using the second modulation.

[0080] The observation point OP21 is a point where the second radar wave is reflected on the right side behind the vehicle VH1 , and the observation point OP22 is a point where the second radar wave is reflected on the right side in front of the vehicle VH1 .

[0081] The observation point OP23 is a point where the second radar wave is reflected on the rear right side of the vehicle VH2 , and the observation point OP24 is a point where the second radar wave is reflected on the front right side of the vehicle VH2 .

[0082] The observation point OP25 is a point where the second radar wave is reflected on the rear right side of the vehicle VH3 , and the observation point OP26 is a point where the second radar wave is reflected on the front right side of the vehicle VH3 .

[0083] Observation points OP27 to OP30 are locations where the second radar wave is reflected by the guardrail GR.

[0084] Through the candidate point clustering process described above, a cluster CL11 including observation points OP21 to OP26 and a cluster CL12 including observation points OP27 to OP30 are generated.

[0085] Cluster CL11 is short along the vehicle's travel direction, so the longitudinal distance extraction condition is not met and it is removed during S320 of the roadside object point cloud extraction process. Furthermore, cluster CL12 is short along the vehicle's travel direction, so the longitudinal distance extraction condition is not met, even though it is at the observation point reflected by the guardrail GR, and it is removed during S320 of the roadside object point cloud extraction process.

[0086] Therefore, when the vertical distance threshold in the vertical distance extraction condition is reduced so that cluster CL12 is set as the "roadside object point cloud", cluster CL11 composed of observation points reflected by vehicles VH1, VH2, and VH3 is also set as the "roadside object point cloud". Figure 8 In FIG. 1 , clusters CL11 and CL12 indicated by solid rectangles are set as “roadside object point clouds”.

[0087] Therefore, in order to set a cluster composed of observation points reflected by the guardrail as a “roadside object point cloud”, it is necessary to use the first modulation instead of the second modulation.

[0088] Next, the steps of the roadside object point cloud association process executed in S30 will be described.

[0089] When performing roadside object point cloud association processing, such as Figure 9 As shown, in S410 , the CPU 11 selects one observation point that has not been selected in the current roadside object point cloud associating process from among the observation points included in the cluster set as the “roadside object point cloud” in S340 (hereinafter referred to as the set cluster).

[0090] In S420 , the CPU 11 calculates the distance between the first modulation observation point selected in S410 (hereinafter, selected observation point) and all second modulation observation points detected in the second modulation period within the same modulation cycle Tm (hereinafter, inter-observation point distance difference).

[0091] For example, in Figure 8 When the observation point OP7 shown is the selected observation point, the CPU 11 calculates the distance between the observation point OP7 and each of the observation points OP21 to OP30 detected by the second modulation as the inter-observation point distance difference.

[0092] In S430, the CPU 11 determines whether a predetermined association condition is satisfied for each of the distance differences between observation points calculated in S420. The association condition in this embodiment is that the distance difference between observation points is the smallest and is lower than a predetermined association threshold.

[0093] Here, if there is no distance difference between the observation points that satisfies the association condition, the CPU 11 moves to S450. On the other hand, if there is a distance difference between the observation points that satisfies the association condition, in S440, the CPU 11 associates the second modulation observation point corresponding to the distance difference between the observation points that satisfies the association condition with the selected observation point (i.e., the first modulation observation point), stores the second modulation observation point information of the second modulation observation point associated with the selected observation point in the RAM 13, and moves to S450. For example, Figure 8 The observation points OP7, OP8, OP9, and OP10 shown are associated with the observation points OP27, OP28, OP29, and OP30, respectively.

[0094] When the process proceeds to S450, the CPU 11 determines whether all first-modulation observation points included in the set cluster have been selected in S410. If not all first-modulation observation points have been selected, the CPU 11 proceeds to S410. On the other hand, if all first-modulation observation points have been selected, the CPU 11 terminates the roadside object point cloud association process.

[0095] Next, the steps of the axis offset angle estimation process executed in S40 will be described.

[0096] When performing the axis offset angle estimation process, such as Figure 10 As shown, in S510, the CPU 11 calculates the distance R, the horizontal azimuth angle θ, and the vertical azimuth angle θ for each of the second modulation observation points (hereinafter, the corresponding observation points) associated with the first modulation observation point in S450 based on the second modulation observation point information of the corresponding observation point. Calculate the position (x, y, z) of the corresponding observation point. The position (x, y, z) is the position in the three-dimensional orthogonal coordinate system with the radar device 2 as the origin. Figure 11 As shown, the X-axis of the three-dimensional orthogonal coordinate system coincides with the central axis CA of radar wave transmission and reception in the radar device 2. The Y-axis of the three-dimensional orthogonal coordinate system is set to extend along the vehicle width direction of the host vehicle VH0 and to be orthogonal to the X-axis. The Z-axis of the three-dimensional orthogonal coordinate system is set to be orthogonal to the X-axis and the Y-axis.

[0097] like Figure 10 As shown, in S520, the CPU 11 determines whether a pre-set deviation determination condition, indicating that the positional deviations of all corresponding observation points are small, is satisfied. The deviation condition determines whether the deviations of the corresponding observation points on the Z-X plane of the three-dimensional orthogonal coordinate system are such that the calculation of the approximate straight line AS, described later, in the subsequent S530 process is difficult. For example, the correlation coefficient of the corresponding observation points on the Z-Y plane can be used as this deviation condition.

[0098] Here, if the deviation determination condition is not satisfied, the CPU 11 ends the axis deviation angle estimation process. On the other hand, if the deviation determination condition is satisfied, in S530 the CPU 11 calculates the approximate straight line AS by the least squares method using the positions (x, y, z) of all corresponding observation points.

[0099] like Figure 11 As shown, the approximate straight line AS is a straight line passing through the ZX plane of the three-dimensional orthogonal coordinate system, and is expressed by the following formula (1): β in the formula (1) is the slope, and C is the intercept. Figure 11 The approximate straight line AS calculated using the positions (x, y, z) of the corresponding observation points OP41 , OP42 , OP43 , OP44 , OP45 , and OP46 reflected by the guardrail GR is shown.

[0100] Z=β·X+C···(1)

[0101] like Figure 10 As shown, in S540 , the CPU 11 calculates the angle corresponding to the slope β of the approximate straight line AS, and calculates the vertical axis offset angle θm by reversing the sign of the angle, and ends the axis offset angle estimation process.

[0102] like Figure 12 As shown, it is assumed that the radar device 2 mounted on the vehicle VH0 transmits and receives the first modulated radar wave (i.e., the first radar wave mentioned above), so that the observation points OP51, OP52, OP53, OP54, OP55, OP56, OP57, OP58, and OP59 reflected on the guardrail GR are detected in order from near to far from the radar device 2.

[0103] In addition, the radar device 2 transmits and receives the second modulated radar wave (i.e., the second radar wave mentioned above), thereby detecting the observation points OP61, OP62, OP63, OP64, OP65, and OP66 reflected by the guardrail GR in order from near to far from the radar device 2.

[0104] As described above, the vertical angle measurement accuracy of the second modulation is higher than that of the first modulation. Therefore, the positional deviation of the observation points OP61 to OP66 of the second modulation is smaller than the positional deviation of the observation points OP61 to OP59 of the first modulation.

[0105] Therefore, the vertical axis offset angle θm calculated by using the second modulated observation points OP61 to OP66 has higher estimation accuracy than the vertical axis offset angle θm calculated by using the first modulated observation points OP61 to OP59.

[0106] The control device 6 configured in this manner estimates the vertical axis deviation angle θm of the radar device 2 mounted on the host vehicle VH0 .

[0107] The radar device 2 is configured to detect a location where the first radar wave is reflected, i.e., a first modulation observation point, by transmitting and receiving a first radar wave modulated by a first modulation method, and repeatedly output a signal including a distance R between the radar device 2 and the first modulation observation point (hereinafter referred to as the first observation point distance), and a horizontal azimuth angle θ and a vertical azimuth angle θ at which the first modulation observation point is located. The first modulated observation point information (hereinafter referred to as the first observation point azimuth).

[0108] The radar device 2 is configured to detect the location where the second radar wave is reflected, i.e., the second modulation observation point, by transmitting and receiving the second radar wave modulated by the second modulation method, and repeatedly output a signal including the distance R between the radar device 2 and the second modulation observation point (hereinafter referred to as the second observation point distance), the horizontal azimuth angle θ at which the second modulation observation point is located, and the vertical azimuth angle θ. The second modulated observation point information (hereinafter referred to as the second observation point azimuth).

[0109] The control device 6 is configured to extract the first modulation observation point reflected by the roadside object from the multiple first modulation observation points detected by the radar device 2 as a roadside object point. The roadside object is arranged on the side of the driving road on which the vehicle VH0 is traveling at a position higher than the driving road and along the direction in which the driving road extends.

[0110] The control device 6 is configured to extract the second modulation observation point corresponding to the roadside object point, i.e., the corresponding observation point, from the plurality of second modulation observation points detected by the radar device 2 based on a pre-set association condition indicating that the position of the roadside object point is close to the position of the second modulation observation point.

[0111] The control device 6 is configured to calculate a vertical axis offset angle θm, in which the center axis CA representing the direction in which the first radar wave and the second radar wave are transmitted and received by the radar device 2, is tilted relative to the front-rear direction of the vehicle VH0 by calculating a slope β representing the direction in which the roadside object extends based on the distribution of the positions of multiple corresponding observation points extracted.

[0112] Such a control device 6 applies a modulation method suitable for detecting roadside objects arranged in the direction extending along the driving road to the first modulation method, and applies a modulation method with higher detection accuracy of the position of the observation point to the second modulation method, so that the slope β representing the direction in which the roadside objects extend can be calculated with better accuracy, thereby improving the estimation accuracy of the vertical axis offset angle θm.

[0113] Furthermore, the maximum detectable distance of the first observation point detected by the first modulation method, i.e., the first maximum detectable distance, is longer than the maximum detectable distance of the second observation point detected by the second modulation method, i.e., the second maximum detectable distance. Furthermore, the accuracy of detecting the azimuth of the second observation point in the vertical direction using the second modulation method, i.e., the second vertical angle measurement accuracy, is higher than the accuracy of detecting the azimuth of the first observation point in the vertical direction using the first modulation method, i.e., the first vertical angle measurement accuracy. Thus, the control device 6 can utilize the first modulation method to detect roadside objects located along the direction in which the travel path extends, and utilize the second modulation method to more accurately calculate the slope β.

[0114] Furthermore, the control device 6 is configured to calculate the vertical axis deviation angle θm by utilizing the fact that the height of the roadside object is constant along the direction in which the travel path extends.

[0115] Furthermore, the control device 6 is configured to calculate the slope β by approximating the distribution of the positions of the plurality of corresponding observation points with a straight line.

[0116] In the embodiment described above, the control device 6 corresponds to the axis deviation estimation device, the host vehicle VH0 corresponds to the mobile object, the vertical axis deviation angle θm corresponds to the axis deviation angle, the first modulation observation point corresponds to the first observation point, and the second modulation observation point corresponds to the second observation point.

[0117] In addition, S10 and S20 are equivalent to the processing of the first roadside object extraction unit, the guardrail GR is equivalent to the roadside object, the roadside object point is equivalent to the first roadside object observation point, and S30 is equivalent to the processing of the second roadside object extraction unit, and the corresponding observation point is equivalent to the second roadside object observation point.

[0118] In addition, S40 corresponds to the processing of the axis deviation angle calculation unit, and the slope β corresponds to the direction information.

[0119] [Second embodiment]

[0120] The following, with the attached Figure 1 The second embodiment of the present disclosure will be described. In the second embodiment, the parts that differ from the first embodiment will be described. The same reference numerals are used for the common structures.

[0121] The axis misalignment detection system 1 of the second embodiment differs from the first embodiment in that the axis misalignment angle estimation process is modified.

[0122] The axis deviation angle estimation process of the second embodiment differs from the first embodiment in that the process of S532 is executed instead of S530 .

[0123] like Figure 13 As shown, when the deviation judgment condition is met in S520, in S532, CPU11 uses the positions of all corresponding observation points (i.e., the observation points of the second modulation) and the positions of all observation points (i.e., the observation points of the first modulation) included in the cluster set as the "roadside object point cloud" in S340 (i.e., the set cluster) that are not associated with the corresponding observation points, calculates the approximate straight line AS by the least squares method, and moves to S540.

[0124] For example, Figure 8 As shown, it is assumed that the radar device 2 detects observation points OP7, OP8, OP9, OP10, OP11, OP12, OP13, OP14, and OP15 reflected by the guardrail GR by transmitting and receiving first modulated radar waves (i.e., the aforementioned first radar waves). Furthermore, it is assumed that the radar device 2 detects observation points OP27, OP28, OP29, and OP30 reflected by the guardrail GR by transmitting and receiving second modulated radar waves (i.e., the aforementioned second radar waves).

[0125] In this case, if Figure 14 As shown, CPU 11 calculates an approximate straight line AS using the least squares method using the positions of observation points OP27, OP28, OP29, and OP30 and the positions of observation points OP11, OP12, OP13, OP14, and OP15. Observation points OP27 to OP30 are corresponding observation points. Observation points OP11 to OP15 are observation points that are not associated with corresponding observation points.

[0126] Control device 6, configured in this manner, is configured to calculate slope β using both the extracted plurality of corresponding observation points and roadside object points not associated with the corresponding observation points among the extracted plurality of roadside object points. This increases the number of observation points used to calculate slope β, thereby further improving the accuracy of estimating vertical axis offset angle θm.

[0127] [Third embodiment]

[0128] The following, with the attached Figure 1 The third embodiment of the present disclosure will be described. In the third embodiment, the parts that differ from the first embodiment will be described. The same reference numerals are used for common structures.

[0129] The axis misalignment detection system 1 of the third embodiment differs from the first embodiment in that the axis misalignment angle estimation process is modified.

[0130] The axis deviation angle estimation process of the third embodiment differs from the first embodiment in that the processes of S514 , S524 , and S534 are executed instead of S510 , S520 , and S530 .

[0131] like Figure 15 As shown, when the axis deviation angle estimation process of the third embodiment is executed, in S514, the CPU 11 estimates the axis deviation angle of each corresponding observation point obtained in the most recent multiple (for example, 5) roadside object point cloud association processes based on the distance R, horizontal azimuth θ and vertical azimuth included in the observation point information of the corresponding observation point. Calculate the position (x, y, z) of the corresponding observation point.

[0132] In S524 , the CPU 11 determines whether a preset deviation determination condition is satisfied, indicating that the deviations among the positions of all corresponding observation points whose positions have been calculated in S514 are small.

[0133] If the deviation determination condition is not met, the CPU 11 terminates the axis deviation angle estimation process. On the other hand, if the deviation determination condition is met, in S534, the CPU 11 calculates an approximate straight line AS using the least squares method using the positions (x, y, z) of all corresponding observation points obtained from the most recent multiple roadside object point cloud association processes, and then proceeds to S540.

[0134] Control device 6, thus configured, is configured to calculate vertical axis offset angle θm by calculating slope β based on the distribution of the positions of a plurality of corresponding observation points obtained over a plurality of modulation periods Tm. This allows control device 6 to increase the number of observation points used to calculate slope β, thereby further improving the accuracy of estimating vertical axis offset angle θm.

[0135] As mentioned above, although one embodiment of the present disclosure has been described, the present disclosure is not limited to the above embodiment, and can be implemented with various modifications.

[0136] [Variation 1]

[0137] While the above embodiment illustrates the transmission and reception of radar waves modulated using the FCM method, any object detection method used by radar device 2 can be used as long as it can detect the position of an object. FCM stands for Fast-Chirp Modulation. For example, an FMCW method or a dual-frequency CW method is also possible. FMCW stands for Frequency Modulated Continuous Wave. Alternatively, a pulse signal transmission and reception method is also possible.

[0138] [Variation 2]

[0139] In the above embodiment, a method of estimating the vertical axis offset angle θm is described, but an axis offset angle in the horizontal direction (ie, a horizontal axis offset angle) may also be estimated.

[0140] [Variation 3]

[0141] In the above embodiment, the association condition of S430 is shown as a method in which the distance difference between the observation points is minimized and the distance difference between the observation points is lower than a preset association threshold. However, the association condition only needs to be a condition indicating that the position of the roadside object point is close to the position of the second modulated observation point. Therefore, the association condition of S430 may also be that the distance difference between the observation points is lower than a preset association threshold. In addition, the association condition of S430 may also be that the difference between the distance R of the roadside object point and the distance R of the second modulated observation point is lower than a preset radius distance threshold, and the difference between the horizontal azimuth angle θ of the roadside object point and the horizontal azimuth angle θ of the second modulated observation point is lower than a preset azimuth angle threshold. The difference between the distance R of the roadside object point and the distance R of the second modulated observation point is equivalent to the radius distance difference, and the difference between the horizontal azimuth angle θ of the roadside object point and the horizontal azimuth angle θ of the second modulated observation point is equivalent to the azimuth angle difference.

[0142] [Variation 4]

[0143] In the above embodiment, the first maximum detection distance is longer than the second maximum detection distance, and the second vertical angle measurement accuracy is higher than the first vertical angle measurement accuracy. However, the first maximum detection distance may be longer than the second maximum detection distance, and the radar device 2 may be configured to not have the function of detecting the azimuth angle of the first observation point in the vertical direction using the first modulation method, but to have the function of detecting the azimuth angle of the second observation point in the vertical direction using the second modulation method. In other words, the control device 6 can estimate the vertical axis offset angle θm even without the vertical angle measurement function using the first modulation method.

[0144] [Variation 5]

[0145] In the above embodiment, a mode in which the control device 6 executes the axis offset adjustment process is shown, but the radar device 2 may also execute the axis offset adjustment process.

[0146] The control device 6 and the method thereof described in the present disclosure may also be implemented by a dedicated computer provided by a processor and a memory programmed to execute one or more functions embodied by a computer program. Alternatively, the control device 6 and the method thereof described in the present disclosure may also be implemented by a dedicated computer provided by a processor constituted by one or more dedicated hardware logic circuits. Alternatively, the control device 6 and the method thereof described in the present disclosure may also be implemented by one or more dedicated computers constituted by a combination of a processor and a memory programmed to execute one or more functions and a processor constituted by one or more hardware logic circuits. In addition, a computer program may also be stored as an instruction executed by a computer in a non-migratable tangible recording medium that can be read by a computer. In the method for implementing the functions of each part included in the control device 6, it is not necessary to include software, and all of its functions may also be implemented using one or more hardware.

[0147] The multiple functions of one component in the above-described embodiment can be realized by multiple components, or the single function of one component can be realized by multiple components. Alternatively, the multiple functions of multiple components can be realized by one component, or the single function realized by multiple components can be realized by one component. Alternatively, a portion of the structure of the above-described embodiment can be omitted. Alternatively, at least a portion of the structure of the above-described embodiment can be added to or replaced with the structure of another above-described embodiment.

[0148] In addition to the above-mentioned control device 6, the present disclosure can also be implemented in various forms such as a system with the control device 6 as a component, a program for causing a computer to function as the control device 6, a non-migratable physical recording medium such as a semiconductor memory recording the program, and an axis offset estimation method.

[0149] [Technical Ideas Disclosed in This Specification]

[0150] [Project 1]

[0151] An axial deviation estimation device is an axial deviation estimation device (6) for estimating an axial deviation angle (θm) of a radar device (2) mounted on a mobile body (VH0), wherein:

[0152] The radar device is configured to detect a first observation point by transmitting and receiving a first radar wave modulated by a first modulation method, and repeatedly output first observation point information including a first observation point distance and a first observation point azimuth, wherein the first observation point is a location at which the first radar wave is reflected, the first observation point distance is the distance between the radar device and the first observation point, and the first observation point azimuth is the azimuth at which the first observation point is located.

[0153] The radar device is configured to detect a second observation point by transmitting and receiving a second radar wave modulated by a second modulation method different from the first modulation method, and repeatedly output second observation point information including a second observation point distance and a second observation point azimuth, wherein the second observation point is a location at which the second radar wave is reflected, the second observation point distance is the distance between the radar device and the second observation point, and the second observation point azimuth is the azimuth at which the second observation point is located.

[0154] The above-mentioned shaft offset estimation device comprises:

[0155] A first roadside object extraction unit (S10, S20) is configured to extract, from the plurality of first observation points detected by the radar device, a first observation point that reflects light from a roadside object as a first roadside object observation point, the roadside object being arranged on the side of a travel path on which the mobile body travels at a position higher than the travel path and in a direction in which the travel path extends;

[0156] a second roadside object extraction unit (S30) configured to extract a second roadside object observation point from the plurality of second observation points detected by the radar device based on a predetermined association condition indicating that the position of the first roadside object observation point is close to the position of the second observation point, the second roadside object observation point being the second observation point corresponding to the first roadside object observation point; and

[0157] The axis deviation angle calculation unit (S40) is configured to calculate the direction information (β) indicating the direction in which the roadside object extends based on the distribution of the positions of the plurality of second roadside object observation points extracted by the second roadside object extraction unit, and thereby calculate the angle at which the central axis (CA) indicating the direction in which the first radar wave and the second radar wave are transmitted and received by the radar device is tilted relative to the front-rear direction of the mobile body as the axis deviation angle.

[0158] [Project 2]

[0159] The shaft offset estimation device according to item 1, wherein:

[0160] The first maximum detection distance is longer than the second maximum detection distance, the first maximum detection distance being the maximum value of the distance between the first observation point detectable by the first modulation method, and the second maximum detection distance being the maximum value of the distance between the second observation point detectable by the second modulation method.

[0161] The second vertical angle measurement accuracy is higher than the first vertical angle measurement accuracy. The second vertical angle measurement accuracy is the accuracy when detecting the azimuth angle of the second observation point in the vertical direction by the second modulation method. The first vertical angle measurement accuracy is the accuracy when detecting the azimuth angle of the first observation point in the vertical direction by the first modulation method.

[0162] [Item 3]

[0163] The shaft offset estimation device according to item 1, wherein:

[0164] The first maximum detection distance is longer than the second maximum detection distance, the first maximum detection distance being the maximum value of the distance between the first observation point detectable by the first modulation method, and the second maximum detection distance being the maximum value of the distance between the second observation point detectable by the second modulation method.

[0165] The radar device is configured not to have a function of detecting the azimuth of the first observation point in the vertical direction using the first modulation method.

[0166] The radar device is configured to have a function of detecting the azimuth angle of the second observation point in the vertical direction using the second modulation method.

[0167] [Item 4]

[0168] The axial offset estimation device according to any one of items 1 to 3, wherein:

[0169] The above-mentioned axis deviation angle calculation unit is configured to calculate the above-mentioned direction information using both the multiple second roadside object observation points extracted by the above-mentioned second roadside object extraction unit and the above-mentioned first roadside object observation points that are not associated with the above-mentioned second roadside object observation points among the multiple first roadside object observation points extracted by the above-mentioned first roadside object extraction unit.

[0170] [Item 5]

[0171] The axial offset estimation device according to any one of items 1 to 3, wherein:

[0172] The radar device is configured to output a plurality of the first observation point information and a plurality of the second observation point information at each preset modulation period.

[0173] The axis deviation angle calculation unit is configured to calculate the axis deviation angle by calculating the direction information based on a distribution of positions of a plurality of second roadside object observation points obtained in a plurality of the modulation cycles.

[0174] [Item 6]

[0175] The axial misalignment estimation device according to any one of items 1 to 5, wherein:

[0176] The association condition is that the distance difference between the observation points is lower than a preset association threshold value for the second observation point, wherein the distance difference between the observation points is the distance between the first roadside object observation point and the second observation point.

[0177] [Item 7]

[0178] The axial misalignment estimation device according to any one of items 1 to 5, wherein:

[0179] The above-mentioned association condition is that the above-mentioned second observation point has the smallest distance difference between the observation points among the multiple above-mentioned second observation points and the distance difference between the above-mentioned observation points is lower than a preset association threshold, wherein the distance difference between the above-mentioned observation points is the distance between the above-mentioned first roadside object observation point and the above-mentioned second observation point.

[0180] [Item 8]

[0181] The axial misalignment estimation device according to any one of items 1 to 5, wherein:

[0182] The above-mentioned association condition is that the radius distance difference is lower than the preset radius distance threshold, and the azimuth angle difference is lower than the preset azimuth angle threshold of the above-mentioned second observation point, wherein the above-mentioned radius distance difference is the difference between the above-mentioned first observation point distance of the above-mentioned first roadside object observation point and the above-mentioned second observation point distance, and the above-mentioned azimuth angle difference is the difference between the above-mentioned first observation point azimuth of the above-mentioned first roadside object observation point and the above-mentioned second observation point azimuth of the above-mentioned second observation point.

[0183] [Item 9]

[0184] The axial misalignment estimation device according to any one of items 1 to 8, wherein:

[0185] The axle misalignment angle calculation unit is configured to calculate the axle misalignment angle by utilizing the fact that the height of the roadside object is constant along the direction in which the travel path extends.

[0186] [Item 10]

[0187] The axial offset estimation device according to any one of items 1 to 9, wherein:

[0188] The axis deviation angle calculation unit is configured to calculate the direction information by approximating a distribution of positions of the plurality of second roadside object observation points using a straight line.

Claims

1. An axial offset estimation device (6) for estimating an axial offset angle (θm) of a radar device (2) mounted on a mobile object (VH0), wherein: The radar device is configured to detect a first observation point by transmitting and receiving a first radar wave modulated by a first modulation method, and repeatedly output first observation point information including a first observation point distance and a first observation point azimuth, wherein the first observation point is a location at which the first radar wave is reflected, the first observation point distance is the distance between the radar device and the first observation point, and the first observation point azimuth is the azimuth at which the first observation point is located. The radar device is configured to detect a second observation point by transmitting and receiving a second radar wave modulated by a second modulation method different from the first modulation method, and repeatedly output second observation point information including a second observation point distance and a second observation point azimuth, wherein the second observation point is a location at which the second radar wave is reflected, the second observation point distance is the distance between the radar device and the second observation point, and the second observation point azimuth is the azimuth at which the second observation point is located. The above-mentioned axis deviation estimation device comprises: A first roadside object extraction unit (S10, S20) is configured to extract, from the plurality of first observation points detected by the radar device, a first observation point that reflects light from a roadside object as a first roadside object observation point, the roadside object being arranged on the side of a travel path on which the mobile body travels at a position higher than the travel path and in a direction in which the travel path extends; a second roadside object extraction unit (S30) configured to extract a second roadside object observation point from the plurality of second observation points detected by the radar device based on a predetermined association condition indicating that the position of the first roadside object observation point is close to the position of the second observation point, the second roadside object observation point being the second observation point corresponding to the first roadside object observation point; and The axis deviation angle calculation unit (S40) is configured to calculate the direction information (β) indicating the direction in which the roadside object extends based on the distribution of the positions of the plurality of second roadside object observation points extracted by the second roadside object extraction unit, and thereby calculate the angle at which the central axis (CA) indicating the direction in which the first radar wave and the second radar wave are transmitted and received by the radar device is tilted relative to the front-rear direction of the mobile body as the axis deviation angle.

2. The axial offset estimation device according to claim 1, wherein: The first maximum detection distance is longer than the second maximum detection distance, the first maximum detection distance being the maximum value of the distance between the first observation point detectable by the first modulation method, and the second maximum detection distance being the maximum value of the distance between the second observation point detectable by the second modulation method. The second vertical angle measurement accuracy is higher than the first vertical angle measurement accuracy. The second vertical angle measurement accuracy is the accuracy when detecting the azimuth angle of the second observation point in the vertical direction by the second modulation method. The first vertical angle measurement accuracy is the accuracy when detecting the azimuth angle of the first observation point in the vertical direction by the first modulation method.

3. The axial offset estimation device according to claim 1, wherein: The first maximum detection distance is longer than the second maximum detection distance, the first maximum detection distance being the maximum value of the distance between the first observation point detectable by the first modulation method, and the second maximum detection distance being the maximum value of the distance between the second observation point detectable by the second modulation method. The radar device is configured not to have a function of detecting the azimuth of the first observation point in the vertical direction using the first modulation method. The radar device is configured to have a function of detecting the azimuth angle of the second observation point in the vertical direction using the second modulation method.

4. The axial offset estimation device according to any one of claims 1 to 3, wherein: The above-mentioned axis deviation angle calculation unit is configured to calculate the above-mentioned direction information using both the multiple second roadside object observation points extracted by the above-mentioned second roadside object extraction unit and the above-mentioned first roadside object observation points that are not associated with the above-mentioned second roadside object observation points among the multiple first roadside object observation points extracted by the above-mentioned first roadside object extraction unit.

5. The axial offset estimation device according to any one of claims 1 to 3, wherein: The radar device is configured to output a plurality of the first observation point information and a plurality of the second observation point information at each preset modulation period. The axis deviation angle calculation unit is configured to calculate the axis deviation angle by calculating the direction information based on a distribution of positions of a plurality of second roadside object observation points obtained in a plurality of the modulation cycles.

6. The axial offset estimation device according to any one of claims 1 to 3, wherein: The association condition is that the distance difference between the observation points is lower than a preset association threshold value for the second observation point, wherein the distance difference between the observation points is the distance between the first roadside object observation point and the second observation point.

7. The axial offset estimation device according to any one of claims 1 to 3, wherein: The above-mentioned association condition is that the above-mentioned second observation point has the smallest distance difference between the observation points among the multiple above-mentioned second observation points and the distance difference between the above-mentioned observation points is lower than a preset association threshold, wherein the distance difference between the above-mentioned observation points is the distance between the above-mentioned first roadside object observation point and the above-mentioned second observation point.

8. The axial offset estimation device according to any one of claims 1 to 3, wherein: The above-mentioned association condition is that the radius distance difference is lower than the preset radius distance threshold, and the azimuth angle difference is lower than the preset azimuth angle threshold of the above-mentioned second observation point, wherein the above-mentioned radius distance difference is the difference between the above-mentioned first observation point distance of the above-mentioned first roadside object observation point and the above-mentioned second observation point distance, and the above-mentioned azimuth angle difference is the difference between the above-mentioned first observation point azimuth of the above-mentioned first roadside object observation point and the above-mentioned second observation point azimuth of the above-mentioned second observation point.

9. The axial offset estimation device according to any one of claims 1 to 3, wherein: The axle misalignment angle calculation unit is configured to calculate the axle misalignment angle by utilizing the fact that the height of the roadside object is constant along the direction in which the travel path extends.

10. The axial offset estimation device according to any one of claims 1 to 3, wherein: The axis deviation angle calculation unit is configured to calculate the direction information by approximating a distribution of positions of the plurality of second roadside object observation points using a straight line.

Images