Information processing method, information processing device, and information processing program

Abstract

This information processing method comprises processing to be implemented by a computer for: acquiring, from a Doppler radar mounted on a mobile body, Doppler velocities of a plurality of targets and azimuth angles of the targets with respect to the mobile body; generating a regression curve indicating the relationship between the Doppler velocity and the azimuth angle on the basis of the acquired Doppler velocities and the orthogonal distances from the azimuth angles to the regression curve if the velocity, the acceleration, and the yaw rate of the mobile body satisfy predetermined conditions; and identifying the velocity of the mobile body on the basis of the generated regression curve.

Description

Information processing method, information processing device, and information processing program

[0001] The present invention relates to an information processing method, an information processing device, and an information processing program for identifying the speed of a moving object.

[0002] In autonomous driving and driving assistance vehicles such as passenger cars and trucks, sensors such as radar perform self-diagnosis of their mounting angles and self-calibration of measured angles, and precise measurement of the vehicle's speed is essential for these processes.In addition, precise vehicle speed is also essential for improving tracking accuracy in tracking processes that continuously identify the positions of surrounding vehicles and pedestrians.

[0003] As such, precise vehicle speed is important measurement data for controlling a vehicle. However, it is known that the vehicle speed sensors used to date have errors due to the vehicle's structure. Vehicle speed sensors generally calculate vehicle speed using the following procedure: A rotary encoder measures tire rotation, and the axle rotation speed is calculated from the number of pulses per time output by the rotary encoder. Vehicle speed is calculated from the axle rotation speed and tire circumference. Therefore, if the tire circumference changes due to insufficient tire pressure or wear, an error will occur between the actual vehicle speed and the calculated vehicle speed.

[0004] In such situations, vehicle speed is detected using a Doppler radar installed on the vehicle. Doppler radar emits microwave or millimeter wave radio waves. The emitted radio waves are called transmitted waves. The transmitted waves are reflected by objects around the vehicle and return. The returning radio waves are called reflected waves. The object that reflects the transmitted waves and is detected by the Doppler radar is called a target. If there is a speed difference between the target and the Doppler radar, the reflected wave will have a frequency shift due to the Doppler effect. This shift is called a Doppler shift. The relative speed to the target can be calculated based on the amount of Doppler shift and the angle indicating the direction in which the reflected wave returned (called the "azimuth angle"). If the target is a stationary object on the road, such as a traffic light, road sign, road lamp, or roadside tree, the calculated relative speed will be equal to the vehicle speed.

[0005] However, since the vehicle is moving between the time when the Doppler radar emits the transmitted wave and the time when the reflected wave is measured, even if the target is a stationary object, a measurement error occurs in the azimuth angle, which results in an error in the vehicle speed. Patent Document 1 proposes a detection device that improves the accuracy of calculated vehicle speed by using only reflected waves whose azimuth angle is smaller than a predetermined threshold, since the smaller the azimuth angle value, the smaller the error.

[0006] Japanese Patent Application Laid-Open No. 2019-7926

[0007] However, in the conventional technology, only targets that return reflected waves with azimuth angles smaller than a predetermined threshold are used, which can make it difficult to stably detect vehicle speed. The present invention has been made in consideration of such circumstances. An object of the present invention is to provide an information processing method, an information processing device, and an information processing program that can accurately and stably determine vehicle speed.

[0008] An information processing method according to one aspect of the present application acquires Doppler velocities of multiple targets and azimuth angles of the targets relative to the moving body from a Doppler radar mounted on the moving body, and if the speed, acceleration, and yaw rate of the moving body satisfy predetermined conditions, generates a regression curve showing the relationship between Doppler velocities and azimuth angles based on the orthogonal distance from the acquired Doppler velocities and azimuth angles to the regression curve, and determines the speed of the moving body based on the generated regression curve, by a computer.

[0009] According to one aspect of the present application, the vehicle speed can be determined accurately and stably.

[0010] FIG. 1 is an explanatory diagram showing an example of the configuration of an information system. FIG. 2 is a block diagram showing an example of the configuration of an ECU and a Doppler radar. FIG. 3 is an explanatory diagram showing the measurement principle of a Doppler radar. FIG. 4 is a graph showing an example of measurement results and a fitting curve of a Doppler radar. FIG. 5 is an explanatory diagram relating to a fitting method. FIG. 6 is an explanatory diagram showing an example of a target list. FIG. 7 is a flowchart showing an example of the procedure of a measurement process. FIG. 8 is a graph showing an example of an estimation result. FIG. 9 is a flowchart showing another example of the procedure of a measurement process. FIG. 10 is a flowchart showing an example of the procedure of a filtering process. FIG. 11 is a flowchart showing another example of the procedure of a filtering process.

[0011] An embodiment will be described below with reference to the drawings. FIG. 1 is an explanatory diagram showing an example of the configuration of an information system. The information system includes an ECU (Electronic Control Unit) 1, a Doppler radar 2, and a vehicle 3. FIG. 1 is a diagram showing the vehicle 3 traveling on a road as viewed from above. The ECU 1 and the Doppler radar 2 are mounted on the vehicle 3. For example, the ECU 1 is installed inside the vehicle cabin. The Doppler radar 2 is installed inside a bumper, emblem, or the like so as not to be visible from the outside. The ECU 1 is an example of an information processing device. The vehicle 3 is an example of a moving object. Alternatively, some information processing may be performed within the Doppler radar, and the remaining information processing may be performed by the ECU 1. The ECU 1 may be configured to be provided inside the Doppler radar 2.

[0012] The target 4 is an object that exists on or near the road. The target includes moving objects such as traffic participants such as automobiles and people, as well as stationary objects on the ground such as obstacles, walls, or telephone poles. The target also includes structures such as road shoulders and sidewalk steps. Moving targets, including those that are temporarily stopped, are called dynamic targets, and stationary targets are called static targets.

[0013] 2 is a block diagram showing an example configuration of an ECU and a Doppler radar. The ECU 1 includes a control unit 11, a storage unit 12, a communication unit 13, and an interface 14. The control unit 11 is configured with a CPU (Central Processing Unit) or an MPU (Micro Processing Unit), etc. The control unit 11 performs various control processes and arithmetic processes by reading and executing a control program 1P and data pre-stored in the storage unit 12. The control unit 11 functions as an acquisition unit, a generation unit, and an identification unit by executing the control program 1P (program product) stored in the storage unit 12.

[0014] The storage unit 12 is configured with memory elements such as RAM (Random Access Memory) or ROM (Read Only Memory), such as DRAM (Dynamic RAM), EEPROM (Electrically Erasable Programmable ROM), or flash memory. The storage unit 12 pre-stores a control program and data referenced during processing. The control program stored in the storage unit 12 may be a control program 1P read from a recording medium 1a readable by the EUC 1. Alternatively, the control program 1P may be downloaded from an external computer (not shown) connected to a communication network (not shown) and stored in the storage unit 12.

[0015] The communication unit 13 is a communication interface that uses a communication protocol such as CAN (Control Area Network), LIN (Local Interconnect Network), or Ethernet (registered trademark), and communicates with in-vehicle devices connected to the in-vehicle network.

[0016] The interface 14 is an interface for serial communication or parallel communication. The interface 14 and the Doppler radar 2 are connected by a cable or the like. The interface 14 is an interface for inputting and outputting data between the control unit 11 and the Doppler radar 2.

[0017] The Doppler radar 2 includes a transmitter 21, a transmitting antenna 22, a receiver 23, and a receiving antenna 24. The transmitter 21 is connected to the ECU 1 via an interface 14. The transmitter 21 transmits radio waves (transmission waves) based on signals from the control unit 11 of the ECU 1. The transmission waves are, for example, radio waves in the millimeter wave band of 30 GHz to 300 GHz or the microwave band of 300 MHz to 300 GHz.

[0018] The transmitting antenna 22 is a directional antenna that transmits radio waves with different strengths depending on the direction. The transmitting antenna 22 is connected to the transmitting unit 21, and transmits radio waves (transmission waves) by output from the transmitting unit 21.

[0019] The receiving antenna 24 is a directional antenna whose reception sensitivity varies depending on the direction. The receiving antenna 24 is connected to the receiving unit 23 and is arranged so as to face in approximately the same direction as the transmitting antenna 22. The receiving antenna 24 receives waves reflected by targets and outputs them to the receiving unit 23. The range in which the Doppler radar 2 can detect targets is determined by the directivity of the transmitting antenna 22 and the directivity of the receiving antenna 24.

[0020] The receiving unit 23 is connected to the ECU 1 via the interface 14. The receiving unit 23 acquires the reflected waves reflected by the target 4 via the receiving antenna 24. The receiving unit 23 performs, for example, A / D conversion on the acquired reflected waves and outputs them to the control unit 11.

[0021] 3 is an explanatory diagram showing the measurement principle of Doppler radar. A transmission wave is emitted from the transmitter 21 of the Doppler radar 2 via the transmission antenna 22. The transmission wave is reflected by targets around the vehicle. The reflected wave is observed by the receiver 23 via the receiving antenna 24. The frequency difference between the transmission wave and the reflected wave is a frequency that corresponds to the relative velocity between the target and the Doppler radar 2. This frequency is called the Doppler frequency. From the Doppler frequency, it is possible to calculate the relative velocity (Doppler velocity) of the target with respect to the Doppler radar 2 attached to the vehicle 3.

[0022] The receiving antenna 24 of the Doppler radar 2 is composed of multiple small antennas with directivity, and the azimuth angle of the reflected wave is calculated from the phase difference of the reflected wave received by each small antenna. The azimuth angle between the target and the Doppler radar 2 is calculated based on the forward direction of the traveling vehicle 3 (front direction of the vehicle). Therefore, if the speed of the vehicle 3 is Host V and the azimuth angle of the target is θ, the relative speed (Doppler velocity) of the target is V d is expressed by the following equation (1).

[0023] V d = -1 * Host V * cosθ … (1)

[0024] When θ is 0, cosθ = 1, and V d = -1 * Host V. That is, the Doppler velocity V dThe extreme value (maximum absolute value) of is equal to the vehicle speed Host V. Doppler velocity V d If the vehicle speed Host V can be measured with high accuracy, the Host V can be calculated with high accuracy.

[0025] However, measurements using the Doppler radar 2 are subject to the following error factors: An installation error when attaching the transmitting antenna 22 and receiving antenna 24 of the Doppler radar 2 to the vehicle 3. Even a slight installation error will cause an error in the azimuth angle θ. Furthermore, in order to avoid spoiling the exterior design of the vehicle 3, the transmitting antenna 22 and receiving antenna 24 are usually housed in a housing and covered with a cover called a radome before being attached to the vehicle 3. When the reflected wave passes through the radome, slight refraction occurs, which will cause an error in the azimuth angle θ.

[0026] Doppler velocity V d and vehicle speed Host V with higher accuracy, the error is corrected. As mentioned above, the azimuth angle θ is used as the explanatory variable, and the Doppler velocity V is used as the objective variable. d The function is a trigonometric function. When there are multiple targets around the vehicle 3 and reflected waves from multiple targets are observed in one measurement, the vehicle speed Host V can be considered to be constant. Therefore, multiple observation data from the Doppler radar 2 are calculated using the azimuth angle θ as an explanatory variable and the Doppler speed V as a target variable. d By fitting to a trigonometric function (cosine function) such that: it is possible to obtain a function in which the error is corrected.

[0027] FIG. 4 is a graph showing an example of Doppler radar measurement results and fitting curves. The horizontal axis is azimuth angle θ, measured in radians, for example. The vertical axis is Doppler velocity, measured in meters per second (m / s) or kilometers per hour (km / h), for example. If the vehicle speed value is positive, the Doppler velocity will be a negative value, but in FIG. 4 it is shown as an absolute value. The points shown in FIG. 4 represent the observation results of multiple targets by the Doppler radar 2. The points showing the observation results will be referred to as observation points hereinafter. The curve is a fitting curve (regression curve) obtained by fitting the observation results.

[0028] 5A and 5B are explanatory diagrams relating to fitting techniques. FIG. 5A is an explanatory diagram of fitting using the least squares method. FIG. 5B is an explanatory diagram of fitting using orthogonal distance regression. First, consider FIG. 5A. As shown in FIG. 5A, in the least squares method, the difference between the observation point and the fitting curve is evaluated along the vertical axis. Therefore, as shown in the left diagram of FIG. 5A, where the slope of the fitting curve is gentle, the difference in the position of the observation point along the horizontal axis is appropriately reflected as the difference along the vertical axis. However, as shown in the right diagram of FIG. 5A, where the slope of the fitting curve is steep, the difference in the position of the observation point along the horizontal axis is excessively reflected as the difference along the vertical axis.

[0029] Next, consider FIG. 5B . As shown in FIG. 5B , in orthogonal distance regression, the difference between the observation points and the fitting curve is evaluated in a direction perpendicular to the fitting curve. Therefore, whether the fitting curve has a gentle slope, as in the left diagram of FIG. 5B , or a steep slope, as in the right diagram of FIG. 5B , the difference in the position of the observation points along the horizontal axis is appropriately reflected as a difference along the vertical axis. From the above, it is clear that orthogonal distance regression is more accurate than the least squares method when fitting observation points to a curve. Therefore, in this embodiment, curve fitting is performed using orthogonal distance regression.

[0030] Next, the data used when performing fitting will be described. FIG. 6 is an explanatory diagram showing an example of a target list. The target list 121 includes an ID sequence, an azimuth sequence, an elevation sequence, a distance sequence, and a Doppler velocity sequence. The ID sequence stores an ID for identifying each target. The ID may simply be a sequential number. The azimuth sequence stores the azimuth angle of the target obtained by observation. The elevation sequence stores the elevation angle of the target obtained by observation. The distance sequence stores the distance to the target obtained by observation. The Doppler velocity sequence stores the Doppler velocity obtained by observation. The target list 121 is created and updated for each observation cycle. Note that the observation results of the Doppler radar 2 are also used for processing related to the vehicle 3 other than the measurement processing described below. The data items shown in FIG. 6 are items particularly necessary for the measurement processing described below, and the target list 121 may include other data items.

[0031] Next, the processing performed by the ECU 1 will be described. Fig. 7 is a flowchart showing an example of the procedure of the measurement processing. The measurement processing is a processing that is repeatedly executed by the ECU 1 while the vehicle 3 is traveling straight ahead at a constant speed. For example, the constant speed means that the vehicle speed is 15 to 120 km / h and the acceleration (component excluding gravitational acceleration) is 1.0 m / s ２ For example, going straight means that the yaw rate is within 1.0 deg / s. The control unit 11 of the ECU 1 acquires the target list 121 (step S1). The control unit 11 determines the driving state of the vehicle 3 (step S2). The driving state is acquired by a process performed by the ECU 1 separately from the measurement process. Depending on the driving state, measurement data suitable for fitting may not be obtained, so this determination is necessary. The control unit 11 determines whether to continue the process (step S3).

[0032] The guideline for the determination is as follows, for example: If the vehicle 3 is traveling straight, the control unit 11 determines that the processing should be continued. If the vehicle 3 is not traveling straight but traveling on a curve, or turning right or left at an intersection or the like, or if the vehicle 3 is traveling within a certain range (for example, ±1.0 m / s ２), the vehicle 3's traveling direction changes, making the state unsuitable for the measurement process. Therefore, the control unit 11 determines not to continue the process. Whether the vehicle 3 is traveling straight or not can be determined by acquiring the vehicle's turning amount and yaw rate from an angular velocity sensor such as a gyro sensor mounted on the vehicle. If the vehicle speed is not within a predetermined range, the control unit 11 determines that the vehicle 3 is in an unsuitable state for the measurement process. This is because if the vehicle speed is below a predetermined first threshold, the Doppler velocity becomes extremely small, making measurement and accurate fitting difficult. Furthermore, if the vehicle speed is above a predetermined second threshold, the angular velocity of the azimuth angle increases, reducing the measurement accuracy of the azimuth angle and making accurate fitting difficult. If the absolute value of the acceleration of the vehicle 3 is smaller than a predetermined third threshold, the control unit 11 determines to continue the measurement process. If the absolute value of the acceleration is not small enough, the Doppler velocity changes during the time period in which the reflected wave is received, making it difficult to measure the Doppler velocity accurately, making it difficult to properly correct the error.

[0033] If the control unit 11 determines not to continue the processing (NO in step S3), it ends the processing for the current cycle. If the control unit 11 determines to continue the processing (YES in step S3), it performs filtering of targets used for fitting (step S4).

[0034] There are four types of filtering. In the first filtering process, only observation points whose azimuth and elevation angles are within a specified range are processed, and other observation points are excluded from processing. In this process, observation points that are not within the normal field of view are excluded from processing.

[0035] In the second filtering process, observation points for which the amount of change between the azimuth angle measured one cycle ago and the azimuth angle measured this time is less than a predetermined value are processed, and observation points for which the amount of change is the predetermined value or more are excluded from processing. In this process, observation points whose positions have changed significantly are excluded from processing. Note that the second filtering process does not necessarily have to be performed.

[0036] In the third filtering process, the error A from the estimated true value of the azimuth angle of the observation point and the error V from the true value of the vehicle speed of the observation point are used to determine the maximum and minimum values ​​of the Doppler velocity error from the maximum and minimum values ​​of the error A and the error V. A range of values ​​that the Doppler velocity should take taking the error into account is determined. If the Doppler velocity of the observation point is within this range, it is retained as a processing target, and if it is not within this range, it is excluded from the processing target. In this process, observation points whose observed Doppler velocity is significantly different from the Doppler velocity expected when the target is assumed to be a static target are excluded.

[0037] The fourth filtering process calculates the Doppler velocity at each observation point. The differences between the Doppler velocity at each observation point and the ideal value (an ideal value without error, expressed as -1 * Host V * cos(θ)) are tallied, and the median (or average) of the differences is calculated. If the difference between each observation point and the median is less than a predetermined value, the observation point is retained as a processing target; if the difference is equal to or greater than the predetermined value, the observation point is excluded from the processing target. In this process, only observation points that are likely to be static targets are retained as processing targets.

[0038] In filtering, the first to fourth filtering processes are performed on all observation points, and the observation points that are determined to be processing targets by all filters are retained. Alternatively, the first filtering process may be performed on all observation points, and the second filtering process may be performed on the observation points that remain as processing targets, and so on.

[0039] The control unit 11 determines whether the number of remaining observation points after filtering is equal to or greater than a predetermined threshold (step S5). If the control unit 11 determines that the number of remaining observation points is less than the threshold (NO in step S5), it ends the processing for the current cycle. If the control unit 11 determines that the number of remaining observation points is equal to or greater than the threshold (YES in step S5), it performs fitting (step S6).

[0040] In the fitting, the running speed of vehicle 3 is Host V, the azimuth angle of the target is θ, and the Doppler velocity V d Then, the fitting curve shown in equation (2) is set.

[0041] V d = -1 * (Host V + α) * cos(θ + β) … (2)

[0042] The control unit 11 uses orthogonal distance regression to find the optimal α and β using the observation points remaining to be processed. The method for finding α and β uses, for example, the BBS algorithm (algorithm of Boggs, Byrd and Schnabel). The BBS algorithm is an algorithm proposed by Paul T. Boggs, Richard H. Byrd, and Robert B. Schnabel. Since it is a well-known algorithm, a detailed explanation will be omitted, but details of the algorithm are disclosed in the following papers, etc.: Paul T. Boggs, Richard H. Byrd, Robert B. Schnabel, "A Stable and Efficient Algorithm for Nonlinear Orthogonal Distance Regression", SIAM Journal on Scientific and Statistical Computing (USA), Society for Industrial and Applied Mathematics, May 1987, Vol. 8, No. 61

[0043] The control unit 11 calculates the vehicle speed and the azimuth angle of the sensor mounting angle based on the fitting result and the observed Doppler velocity (step S7). The vehicle speed is an extreme value in the fitting curve. The azimuth angle of the sensor mounting angle is an azimuth angle that indicates an extreme value. The control unit 11 stores the calculated vehicle speed and the azimuth angle of each target (step S8), and ends the process. As described above, the measurement process shown in FIG. 7 is a periodic process that is periodically repeated.

[0044] FIG. 8 is a graph showing an example of an estimation result. The horizontal axis represents time in seconds. The vertical axis represents vehicle speed in meters per second. As shown in FIG. 8, compared to the graph of the prior art, the graph of this embodiment has fewer large fluctuations in values, and it can be seen that accuracy is improved.

[0045] This embodiment has the following advantages: Since the orthogonal distance regression is used as the fitting method, rather than the conventionally widely used least squares method, it is possible to obtain highly accurate results. As a result, it is possible to estimate the vehicle speed with high accuracy.

[0046] Since the vehicle speed can be estimated with high accuracy, it is possible to estimate and correct the azimuth angle measurement error using the vehicle speed. That is, using the above equation (1), θ= acos(V d In addition, if the Doppler radar 2 is mounted at an angle of 62 degrees counterclockwise from the front of the vehicle instead of the original angle of 60 degrees, an error of +2 degrees can be detected simultaneously with the vehicle speed.

[0047] In the above description, the vehicle 3 is equipped with only the ECU 1, but this is not limited to this. The vehicle 3 may be equipped with multiple ECUs. When multiple ECUs are provided, it is common to provide a main ECU that controls the entire vehicle, and multiple sub-ECUs that control various parts of the vehicle under the control of the main ECU. In this case, one of the sub-ECUs is the above-mentioned ECU 1, and is named, for example, a radar ECU.

[0048] In the above description, the ECU 1 and the Doppler radar 2 are separate entities, but this is not limiting. In particular, when a sub-ECU including a main ECU and a radar ECU is provided, the radar ECU and the Doppler radar may be integrated into a single device.

[0049] When the above-described information system is configured using a CCU (Central Computing Unit), the above-described ECU (Electronic Control Unit) 1 and storage medium 1a are provided in the CCU. The Doppler radar 2 may be provided externally as a satellite radar. Furthermore, the above-described reference data may be data previously generated by the ECU 1 or the CCU itself.

[0050] In the above description, the transmitting antenna 22 and the receiving antenna 24 are each a directional antenna, but this is not limiting. Non-directional antennas or omnidirectional antennas may be used as the transmitting antenna 22 and the receiving antenna 24 as long as they can perform the functions required of the transmitting and receiving antennas of the Doppler radar 2.

[0051] Next, another embodiment of the processing performed by the ECU 1 will be described. Fig. 9 is a flowchart showing another example of the procedure of the measurement processing. The measurement processing is a processing that is repeatedly executed by the ECU 1 while the vehicle 3 is traveling straight forward at a constant speed. For example, the constant speed means that the vehicle speed is 15 to 120 km / h and the acceleration (component excluding gravitational acceleration) is 1.0 m / s ２ For example, going straight means that the yaw rate is within 1.0 deg / s. The control unit 11 of the ECU 1 acquires the target list 121 (step S11). The control unit 11 determines the driving state of the vehicle 3 (step S12). The driving state is acquired by a process performed by the ECU 1 separately from the measurement process. Depending on the driving state, measurement data suitable for fitting may not be obtained, so the control unit 11 determines the driving state. The control unit 11 determines whether or not to continue processing based on the result of the driving state determination (step S13).

[0052] The guideline for determining whether or not the process needs to be continued is, for example, as follows: If the vehicle 3 is traveling straight, the control unit 11 determines that the process should be continued. If the vehicle 3 is not traveling straight but traveling on a curve, or turning right or left at an intersection, or if the vehicle 3 is traveling within a certain range (for example, ±1.0 m / s ２), the vehicle 3 is in an unsuitable state for the measurement process because the vehicle 3's traveling direction changes or the amount of change in vehicle speed during Doppler velocity sampling increases. Therefore, the control unit 11 determines not to continue the process. Whether the vehicle 3 is traveling straight or not can be determined by acquiring the vehicle's turning amount and yaw rate from an angular velocity sensor such as a gyro sensor mounted on the vehicle. If the vehicle speed is not within a predetermined range, the control unit 11 determines that the vehicle 3 is in an unsuitable state for the measurement process. This is because if the vehicle speed is below a predetermined first threshold (e.g., 15 km / h), the Doppler velocity becomes extremely small, making measurement and accurate fitting difficult. Furthermore, if the vehicle speed is above a predetermined second threshold (e.g., 120 km / h), the angular velocity of the azimuth angle increases, reducing the accuracy of azimuth angle measurement and making accurate fitting difficult. The control unit 11 determines whether the absolute value of the acceleration of the vehicle 3 is below a predetermined third threshold (e.g., ±1.0 m / s ２ ), it is determined that the measurement process should be continued. If the absolute value of the acceleration is not small enough, the Doppler velocity changes during the time period in which the reflected wave is received, making it difficult to measure the Doppler velocity accurately, and making it difficult to appropriately correct the error.

[0053] If the control unit 11 determines not to continue the processing (NO in step S13), it ends the processing for the current cycle. If the control unit 11 determines to continue the processing (YES in step S13), it performs filtering of targets used for fitting (step S14). For example, in this embodiment, five types of filtering processing are performed.

[0054] 10 is a flowchart showing an example of the filtering process procedure. The control unit 11 performs a first filtering process (step S31). In the first filtering process, only observation points whose azimuth angle and elevation angle are within a predetermined range are processed. For example, it is determined whether the predetermined range of azimuth angle is + / - 75 degrees from the front of the Doppler radar, and the predetermined range of elevation angle is + / - 20 degrees.

[0055] The control unit 11 performs filtering (step S32). The control unit 11 retains observation points whose azimuth and elevation angles satisfy the conditions as processing targets, and excludes those that do not. Through the first filtering process, the control unit 11 excludes observation points that are not within the normal field of view from the processing targets.

[0056] The control unit 11 performs a second filtering process (step S33). In the second filtering process, observation points for which the amount of change between the azimuth angle measured one cycle ago and the azimuth angle measured this time is less than a predetermined value are processed. For example, it is determined whether the amount of change between the azimuth angle of the observation point measured one cycle ago and the azimuth angle measured this time is within ±20°.

[0057] Control unit 11 performs filtering (step S34). Control unit 11 keeps as processing targets those observation points whose amount of change in azimuth angle satisfies the conditions, and excludes from processing those that do not. In the second filtering process, control unit 11 excludes from processing targets observation points whose positions have changed significantly through observation point tracking processing.

[0058] The control unit 11 then performs a third filtering process (step S35). In this process, the range of error in the observation is taken into consideration, and observation points that are clearly moving relative to the ground are removed, and all observation points that may be stationary are processed. For example, it is determined whether the observation point is within the range of maximum and minimum Doppler values ​​(predetermined value range) that is calculated using a range of ±10° for the azimuth angle of the observation point and a range of ±15% for the true value of the vehicle speed.

[0059] The control unit 11 performs screening (step S36). The control unit 11 leaves as processing targets observation points that are observed within a predetermined range, and excludes from processing targets those that are not.

[0060] In the fourth filtering process (step S37), the control unit 11 selects observation points that are likely to be static targets, taking into account the influence of errors in the vehicle speed. For example, for each observation point, assuming that the vehicle speed and azimuth angle are accurate, the control unit 11 calculates the ideal value of the Doppler velocity (vehicle speed * cos(azimuth angle)), compares the observed Doppler velocity with the ideal value, and determines the vehicle speed error (vehicle speed error = (observed Doppler velocity value / ideal Doppler velocity value) - 1). The control unit 11 selects the median (or average) of the obtained vehicle speed errors and sets this value as the provisional vehicle speed error. Using this provisional vehicle speed error, the control unit 11 recalculates the expected Doppler velocity for each observation point ((1 + vehicle speed error) * vehicle speed * cos(azimuth angle)). If the observed Doppler velocity is within a certain range, the fourth filtering process is terminated. The control unit 11 then returns the process to the caller.

[0061] The control unit 11 performs screening (step S38), leaving only observation points whose observed Doppler velocities are within a certain range, and excluding those that do not.

[0062] The control unit 11 performs a fifth filtering process, which is a process of thinning out the observation points in order to improve the accuracy of fitting (step S39). For example, the control unit 11 divides the observation range of the azimuth angle into multiple sections each of which is 1 degree, and from each section (the same section), a maximum of one (predetermined number) of observation points is left as a processing target as a threshold value, and the second and subsequent observation points are excluded from the processing target.

[0063] The control unit 11 performs filtering (step S40). For each observation point, only the observation points that remain after the fifth filtering are passed on to the next processing step, and the remaining observation points are not passed on to the next processing step.

[0064] In the filtering process shown in FIG. 10, the first filtering process is performed on all observation points, and then the second filtering process is performed on the observation points that remain to be processed, and so on, so that filtering is performed sequentially.

[0065] 11 is a flowchart showing another example of the procedure for filtering processing. Control unit 11 performs a first filtering process for all observation points (step S51). The processing content is as described above. Control unit 11 stores whether or not each observation point was determined to be a processing target in the first filtering process (step S52).

[0066] The control unit 11 performs the second filtering process for all observation points (step S53). The processing content is as described above. The control unit 11 stores whether or not each observation point was determined to be a processing target in the second filtering process (step S54).

[0067] Control unit 11 performs the third filtering process on all observation points (step S55). The process is as described above. Control unit 11 stores whether or not each observation point has been determined to be a processing target in the third filtering process (step S56).

[0068] Control unit 11 performs the fourth filtering process on all observation points (step S57). The processing content is as described above. Control unit 11 stores whether or not each observation point was determined to be a processing target in the fourth filtering process (step S58).

[0069] Control unit 11 performs a fifth filtering process on all observation points (step S59). The process is as described above. Control unit 11 stores whether or not each observation point has been determined to be a processing target in the fifth filtering process (step S60).

[0070] The control unit 11 narrows down the observation points (step S61). The control unit 11 leaves, from among all the observation points, the observation points that have been determined to be processing targets in all filtering processes. The control unit 11 returns the process to the caller.

[0071] The control unit 11 may perform the first to fifth filtering processes (steps S51, S53, S55, S57, and S59) and the subsequent storage processes (steps S52, S54, S56, S58, and S60) in parallel. In this case, the control unit 11 executes step S61 after all filtering processes and storage processes are completed.

[0072] There are no limitations on the order in which the five filtering processes, the first through fifth, are executed or the number of filtering processes to be selected. It is not necessary to execute all five of these filtering processes; any necessary filtering processes may be implemented as appropriate. Furthermore, filtering processes other than the exemplified filtering processes may be added. For example, the first filtering process, the third filtering process, the fourth filtering process, and the fifth filtering process may be implemented and executed in this order. Furthermore, the fourth filtering process may be omitted if it is confirmed in the third process, the tracking process, or the selection process that most or all of the observation points are likely to be stationary relative to the ground.

[0073] The control unit 11 determines whether the observation points remaining after the filtering process are observed within a predetermined range within the observation range of the azimuth angle of the Doppler radar (step S15). For example, the control unit 11 determines that the remaining observation points are observed within the predetermined range if the maximum and minimum values ​​of the azimuth angle of the remaining observation points are within a range of 50% or more of the observation range (field of view) of the azimuth angle of the Doppler radar. "Existing within a range of 50% or more" means that, when the Doppler radar has an observation range of ±θ with respect to the azimuth angle, i.e., 2θ, the difference between the maximum and minimum values ​​of the azimuth angle of the remaining observation points is θ or more. If the control unit 11 determines that the observation points are not observed within the predetermined range (NO in step S15), it ends the processing for the current cycle.

[0074] If the control unit 11 determines that the observation points are within the predetermined range (YES in step S15), the control unit 11 performs fitting using the observation points that remain to be processed in the above processing (step S16). The fitting is the same as the content of step S4 described above, so a description thereof will be omitted.

[0075] The control unit 11 calculates the vehicle speed and the azimuth angle of the sensor mounting angle based on the fitting result and the observed Doppler velocity (step S17). The content of step S17 is the same as step S7 described above, so a description thereof will be omitted. The control unit 11 stores the calculated vehicle speed and the azimuth angle of each target (step S18), and ends the process.

[0076] The graph of the estimation results in this embodiment, like the graph of FIG. 8 referred to above, has improved accuracy compared to the graph of the prior art.

[0077] The technical features (constituent elements) described in each embodiment can be combined with each other, and by combining them, new technical features can be formed. The embodiments disclosed herein are illustrative in all respects and should not be considered limiting. The scope of the present invention is defined by the claims, not the meanings described above, and is intended to include all modifications within the meaning and scope of the claims. Furthermore, independent and dependent claims described in the claims can be combined with each other in any and all combinations, regardless of the reference format. Furthermore, although the claims use a format in which a claim references two or more other claims (multiple claim format), this is not limited to this format. A multiple claim (multi-multi claim) that references at least one other multiple claim may also be used.

[0078] 1: ECU 11: Control unit 12: Storage unit 121: Target list 13: Communication unit 14: Interface 1P: Control program 1a: Recording medium 2: Doppler radar 21: Transmitter 22: Transmitting antenna 23: Receiving unit 24: Receiving antenna 3: Vehicle 4: Target

Claims

1. An information processing method in which a computer performs the following processes: acquiring Doppler velocities of multiple targets and azimuth angles of the targets relative to the moving body from a Doppler radar mounted on a moving body; if the speed, acceleration, and yaw rate of the moving body satisfy predetermined conditions, generating a regression curve showing the relationship between Doppler velocity and azimuth angle based on the orthogonal distance from the acquired Doppler velocity and azimuth angle to the regression curve; and determining the speed of the moving body based on the generated regression curve.

2. The information processing method according to claim 1, further comprising the step of calculating the azimuth angle of the target based on the velocity and the Doppler velocity of the target.

3. The information processing method according to claim 1 or 2, wherein the azimuth angle is repeatedly output by periodic processing.

4. The information processing method according to claim 3, further comprising the steps of: performing a filtering process including judging whether the target can be regarded as a static target; and using the target that has passed the filtering process as a processing object for generating the regression curve.

5. The information processing method according to claim 4, wherein the filtering process includes: a first filtering process for passing the acquired azimuth angle if it is within a predetermined range; a second filtering process for calculating a Doppler velocity of the target object and passing the Doppler velocity if the calculated Doppler velocity is within a range of values ​​that should be taken with an error taken into account; and a third filtering process for tallying up the differences between the Doppler velocity of each of the multiple targets and an ideal value of the Doppler velocity, calculating a median or average of the differences, recalculating the Doppler velocity by assuming that the calculated median or average is an error, and passing the recalculated Doppler velocity if it is within a predetermined value range.

6. The information processing method according to claim 5, further comprising the steps of: performing the second filtering process on targets that have passed the first filtering process; and performing the third filtering process on targets that have passed the second filtering process.

7. The information processing method according to claim 5, wherein the first filtering process, the second filtering process, and the third filtering process are performed in parallel for the same target, and only targets that have passed all of the filtering processes are processed.

8. The information processing method according to claim 4, further comprising the steps of: generating a regression curve for targets that have passed the filtering process, if the number of targets whose azimuth angles fall within a predetermined range is equal to or greater than a threshold; and determining the speed of the moving object based on the generated regression curve.

9. The information processing method according to claim 8, further comprising the steps of: dividing an observable azimuth angle range into a plurality of intervals; determining in which interval a target object that has passed the filtering process is included; if a predetermined number of targets are included in the same interval, generating a regression curve for the predetermined number of targets as processing objects; and determining the speed of the moving object based on the generated regression curve.

10. The information processing method according to claim 4, wherein the filtering process includes: a first filtering process for passing the target when the azimuth angle of the target falls within a predetermined range; a third filtering process for removing observation points where the target is clearly moving relative to the ground and passing all static targets which are likely to be stationary; and a fourth filtering process for removing observation points which are unlikely to be static targets, taking into account the influence of errors in the vehicle speed of the moving body, and passing observation points which are likely to be static targets.

11. The information processing method according to claim 1 or 2, wherein the regression curve is determined using a BBS algorithm.

12. An information processing device comprising: an acquisition unit that acquires Doppler velocities of multiple targets and azimuth angles of the targets relative to the moving body from a Doppler radar mounted on the moving body; a generation unit that generates a regression curve showing the relationship between Doppler velocity and azimuth angle based on the orthogonal distance from the acquired Doppler velocity and azimuth angle to the regression curve when the velocity, acceleration, and yaw rate of the moving body satisfy predetermined conditions; and an identification unit that identifies the velocity of the moving body based on the generated regression curve.

13. An information processing program that causes a computer to perform the following processes: acquiring Doppler velocities of multiple targets and azimuth angles of the targets relative to the moving body from a Doppler radar mounted on a moving body; if the speed, acceleration, and yaw rate of the moving body satisfy predetermined conditions, generating a regression curve showing the relationship between Doppler velocity and azimuth angle based on the orthogonal distance from the acquired Doppler velocity and azimuth angle to the regression curve; and determining the speed of the moving body based on the generated regression curve.

Images