Steering control device

Abstract

To provide a steering control device that is able to calculate a learning value for correcting a target lateral position according to the magnitude of influence on an error in the vehicle lateral position caused by deviation of a mounting position.SOLUTION: A steering control device 100 executes steering control including steering assistance or automatic steering of a vehicle 20. The steering control device 100 includes: a path generation unit 14 that generates a target path including a target lateral position, based on a captured image showing ahead of the vehicle 20; a learning-value calculation unit 15 that calculates a learning value for correcting the target lateral position, based on an amount of deviation from the target lateral position that a manual path of the vehicle 20 following a manual steering operation has with respect to the target path when a driver of the vehicle 20 performs the manual steering operation during the execution of the steering control; and a vehicle control unit 16 that executes the steering control, based on the target path and the learning value. The learning-value calculation unit 15 calculates a first learning-value, which is a learning value when the vehicle 20 is traveling on a straight road; and a second learning-value, which is a learning value when the vehicle 20 is traveling on a curved road.SELECTED DRAWING: Figure 1

Description

[Technical Field]

[0001] The present disclosure relates to a steering control device. [Background technology]

[0002] In the vehicle position estimation device described in Patent Document 1, when a steering override is determined, the vehicle lateral position control is suspended, and when the steering override is subsequently determined to have ended, the vehicle lateral position control is resumed. The difference between the vehicle position based on GPS information and the vehicle position based on map information is regarded as an inherent position error of the GPS information, and this inherent position error is reflected in the vehicle position at the end of the steering override. [Prior art documents] [Patent documents]

[0003] [Patent Document 1] Japanese Patent Application Publication No. 2017-013586 Summary of the Invention [Problem to be solved by the invention]

[0004] In the past, the mounting orientation of a camera used for steering control could deviate from the ideal state, for example, if there was an assembly error during shipping or if the user unintentionally touched the camera and shifted its orientation. The effect of such a deviation in mounting orientation on the error in the vehicle's lateral position is not necessarily uniform between straight and curved roads. Therefore, there was room for improvement in error correction. [Means for solving the problem]

[0005] One aspect of the present disclosure is a steering control device that executes steering control including steering assistance or automatic steering of a vehicle, and includes: a path generation unit that generates a target path including a target lateral position in the vehicle's driving lane based on a captured image of a road ahead of the vehicle; a learning value calculation unit that, when a driver of the vehicle performs a manual steering operation while the steering control is being executed, calculates a learning value to correct the target lateral position based on an amount of deviation of the manual path of the vehicle in accordance with the manual steering operation from the target lateral position with respect to the target path; and a steering control unit that executes steering control based on the target path and the learning value, wherein the learning value calculation unit calculates a first learning value that is a learning value when the vehicle is traveling on a straight road, and a second learning value that is a learning value when the vehicle is traveling on a curved road.

[0006] In a steering control device according to one aspect of the present disclosure, a first learned value for a straight road and a second learned value for a curved road are calculated as learned values ​​for correcting a target lateral position. An error in the vehicle's lateral position on a road on which the vehicle is traveling may be caused by a deviation in the mounting orientation of a camera that captures captured images. The error in the vehicle's lateral position caused by a deviation in the mounting orientation is not necessarily uniform between a straight road and a curved road; for example, the error may be larger on a curved road than on a straight road. According to the above configuration, the first learned value for a straight road and the second learned value for a curved road may be calculated as different values. Therefore, a learned value for correcting the target lateral position can be calculated depending on the magnitude of the influence on the error in the vehicle's lateral position caused by a deviation in the mounting orientation.

[0007] In one embodiment, the learning value calculation unit may calculate a right learning value that is the second learning value when the vehicle is traveling on a right-hand curve and a left learning value that is the second learning value when the vehicle is traveling on a left-hand curve. The effect of a deviation in the mounting posture on the error in the vehicle lateral position is not necessarily uniform when the vehicle is traveling on a right-hand curve and when the vehicle is traveling on a left-hand curve. According to the above configuration, the right learning value for a right-hand curve and the left learning value for a left-hand curve can be calculated as different values. Therefore, the second learning values ​​can be calculated corresponding to the difference in the magnitude of the effect of a deviation in the mounting posture on the error in the vehicle lateral position caused by the deviation between a right-hand curve and a left-hand curve.

[0008] In one embodiment, the learning value calculation unit may calculate the learning value so that the learning speed of the second learning value is faster than the learning speed of the first learning value. In this case, by making the learning speed faster when traveling on a curved road than when traveling on a straight road, the vehicle lateral position can be corrected earlier on a curved road, which tends to have a large effect on an error in the vehicle lateral position. [Effects of the Invention]

[0009] According to various aspects of the present disclosure, it is possible to calculate a learning value that corrects the target lateral position depending on the magnitude of the effect on the error in the vehicle lateral position caused by a deviation in the mounting attitude. [Brief explanation of the drawings]

[0010] [Figure 1] 1 is a schematic configuration diagram of an example of a vehicle including a steering control device according to an embodiment; [Figure 2] FIG. 10 is a plan view for explaining calculation of a learning value for correcting a target lateral position based on the amount of deviation from the target lateral position. [Figure 3] 10 is a flowchart illustrating an example of a process for calculating a learning value. [Figure 4] 4 is a flowchart showing an example of a calculation process of a second learned value in FIG. 3. [Figure 5] 10 is a flowchart illustrating an example of a learning rate setting process. DETAILED DESCRIPTION OF THE INVENTION

[0011] Hereinafter, exemplary embodiments will be described with reference to the drawings. In the following description, the same or equivalent elements will be denoted by the same reference numerals, and redundant description may be omitted.

[0012] FIG. 1 is a schematic configuration diagram of an example of a vehicle including a steering control device according to an embodiment. The steering control device 100 shown in FIG. 1 is an automatic driving device mounted on a vehicle 20 such as a passenger car. The steering control device 100 is configured to be able to execute steering control including steering assistance for the vehicle 20. The steering assistance is a driving state in which control is performed to assist the driver in steering the vehicle 20. The steering assistance is, for example, lane tracing control (LKA [Lane Keeping Assist]) that prompts the driver to steer the vehicle 20 so as not to deviate from the driving lane.

[0013] The steering control device 100 may be configured to be able to execute steering control including automatic steering of the vehicle 20. Automatic steering is a driving state in which the steering of the vehicle 20 is automatically controlled. Automatic steering may, for example, be performed by controlling the LKA so as not to deviate from the driving lane, without the driver operating the steering of the vehicle 20.

[0014] The steering control device 100 may be configured to be capable of performing automatic driving including automatic steering. Automatic driving is vehicle control that automatically drives the vehicle 20 along a predetermined target route. The target route here is a route on a map along which the vehicle 20 travels under automatic driving control. In automatic driving, the driver does not need to perform driving operations such as steering and acceleration / deceleration, and the vehicle 20 travels automatically.

[0015] In other words, the steering control of this embodiment may refer to control of the LKA, including at least one of steering assistance, automatic steering not during automatic driving, and automatic steering during automatic driving. Of the above-mentioned LKA controls, steering assistance and automatic steering not during automatic driving can be performed based on, for example, an image of the area ahead of the vehicle 20 captured by an on-board camera. Each of the above-mentioned LKA controls may also be performed based on map information.

[0016] [Configuration of steering control device 100] As shown in Fig. 1, the steering control device 100 includes an ECU (Electronic Control Unit) 10 that manages steering control. The ECU 10 is an electronic control unit that includes a CPU (Central Processing Unit), a ROM (Read Only Memory), a RAM (Random Access Memory), a CAN (Controller Area Network) communication circuit, and the like. The ECU 10, for example, loads a program stored in the ROM into the RAM and executes the program loaded into the RAM with the CPU, thereby realizing various functions. The ECU 10 may be composed of multiple electronic control units.

[0017] The ECU 10 is connected to a GPS (Global Positioning System) receiver 1 , an external sensor 2 , an internal sensor 3 , a map database 4 , and an actuator 5 .

[0018] The GPS receiver 1 receives signals from three or more GPS satellites to measure the position of the vehicle 20 (for example, the latitude and longitude of the vehicle 20). The GPS receiver 1 transmits the measured position information of the vehicle 20 to the ECU 10.

[0019] The external sensor 2 is a detector that detects the situation around the vehicle 20. The external sensor 2 includes at least a camera. The external sensor 2 may also include a radar sensor.

[0020] The camera is an imaging device that captures images of the external situation of the vehicle 20. The camera is provided behind the windshield of the vehicle 20. The camera transmits captured images of the external situation including the area ahead of the vehicle 20 to the ECU 10. The camera may be a monocular camera or a stereo camera.

[0021] A radar sensor is a detector that detects obstacles around the vehicle 20 using radio waves (e.g., millimeter waves) or light. Radar sensors include, for example, millimeter wave radar or LIDAR (Light Detection and Ranging). The radar sensor detects obstacles by transmitting radio waves or light to the periphery of the vehicle 20 and receiving the radio waves or light reflected by the obstacles. The radar sensor transmits detected obstacle information to the ECU 10. Obstacles include fixed obstacles that define road lanes, such as curbs, guardrails, poles, or safety cones. Obstacles may also include moving obstacles such as pedestrians, bicycles, and other vehicles.

[0022] The internal sensor 3 is a detector that detects the traveling state of the vehicle 20. The internal sensor 3 includes a vehicle speed sensor, an acceleration sensor, and a yaw rate sensor. The vehicle speed sensor is a detector that detects the speed of the vehicle 20. As the vehicle speed sensor, for example, a wheel speed sensor that is provided on a wheel of the vehicle 20 or a drive shaft that rotates integrally with the wheel and detects the rotation speed of the wheel is used. The vehicle speed sensor transmits the detected vehicle speed information (wheel speed information) to the ECU 10.

[0023] The acceleration sensor is a detector that detects the acceleration of the vehicle 20. The acceleration sensor includes, for example, a longitudinal acceleration sensor that detects the acceleration in the longitudinal direction of the vehicle 20 and a lateral acceleration sensor that detects the lateral acceleration of the vehicle 20. The acceleration sensor transmits, for example, acceleration information of the vehicle 20 to the ECU 10. The yaw rate sensor is a detector that detects the yaw rate (rotational angular velocity) around the vertical axis of the center of gravity of the vehicle 20. A gyro sensor, for example, can be used as the yaw rate sensor. The yaw rate sensor transmits the detected yaw rate information of the vehicle 20 to the ECU 10.

[0024] The map database 4 is a database that stores map information. The map database 4 is formed, for example, in a storage device such as a hard disk drive (HDD) mounted on the vehicle 20. The map information includes road position information, road shape information (for example, curves, types of straight sections, curvature of curves, information on whether the curve is a right-hand curve or a left-hand curve, etc.), and position information of intersections and branching points. Note that the map database 4 is not essential unless the vehicle 20 is configured to be able to perform autonomous driving.

[0025] The actuator 5 is a device used to control the vehicle 20. The actuator 5 includes at least a steering actuator. The steering actuator controls the drive of an assist motor that controls the steering torque of the electric power steering system in accordance with a control signal from the ECU 10. In this way, the steering actuator controls the steering torque of the vehicle 20.

[0026] The actuator 5 may include a drive actuator and a brake actuator. The drive actuator controls the amount of air supplied to the engine (throttle opening) in response to a control signal from the ECU 10, thereby controlling the driving force of the vehicle 20. If the vehicle 20 is a hybrid vehicle, in addition to the amount of air supplied to the engine, a control signal from the ECU 10 is input to a motor serving as a power source to control the driving force. If the vehicle 20 is an electric vehicle, a control signal from the ECU 10 is input to a motor serving as a power source to control the driving force. In these cases, the motor serving as a power source constitutes the actuator 5. The brake actuator controls a brake system in response to a control signal from the ECU 10, thereby controlling the braking force applied to the wheels of the vehicle 20. As the brake system, for example, a hydraulic brake system can be used.

[0027] Next, a description will be given of the functional configuration of the ECU 10. The ECU 10 has a surrounding environment recognition unit 11, a vehicle position recognition unit 12, a driving state recognition unit 13, a route generation unit 14, a learning value calculation unit 15, and a vehicle control unit (steering control unit) 16. Some of the functions of the ECU 10 may be executed by a server that can communicate with the vehicle 20.

[0028] The surrounding environment recognition unit 11 recognizes the surrounding environment of the vehicle 20 based on the detection results of the external sensor 2. The surrounding environment includes the status of lane markings (e.g., white lines) around the vehicle 20. The surrounding environment may also include the status of obstacles (including curbs) around the vehicle 20.

[0029] The surrounding environment recognition unit 11 acquires a captured image of the area ahead of the vehicle 20 using the camera of the external sensor 2. The surrounding environment recognition unit 11 recognizes dividing lines around the vehicle 20 based on the captured image of the area ahead of the vehicle 20, and recognizes the driving lane in which the vehicle 20 is traveling. The surrounding environment recognition unit 11 may also acquire obstacle information such as a curb in front of the vehicle 20 using a radar sensor of the external sensor 2, and recognize the driving lane in which the vehicle 20 is traveling.

[0030] The vehicle position recognition unit 12 recognizes the position of the vehicle 20 on the driving lane or the position of the vehicle 20 on a map. The vehicle position recognition unit 12 recognizes the relative position of the vehicle 20 and a lane marking based on, for example, an image of the front of the vehicle 20 captured by a camera of the external sensor 2, and recognizes the position of the vehicle 20 on the driving lane. The position of the vehicle 20 on the driving lane may be obtained, for example, as coordinates corresponding to the midpoint between a pair of left and right drive wheels of the vehicle 20.

[0031] If the vehicle 20 is configured to be able to perform automatic steering during automatic driving, the vehicle position recognition unit 12 may recognize the position of the vehicle 20 on a map based on the position information of the GPS receiver 1 and the map information of the map database 4. The vehicle position recognition unit 12 may use the position information of targets included in the map information of the map database 4 and the detection results of the external sensor 2 to accurately recognize the position of the vehicle 20 using SLAM (Simultaneous Localization and Mapping) technology or the like. The vehicle position recognition unit 12 may also recognize the position of the vehicle 20 on a map using other well-known methods.

[0032] The running state recognition unit 13 recognizes the running state of the vehicle 20 based on the detection results of the internal sensor 3. The running state includes the speed of the vehicle 20, the acceleration of the vehicle 20, and the yaw rate of the vehicle 20. Specifically, the running state recognition unit 13 recognizes the speed of the vehicle 20 based on vehicle speed information from a vehicle speed sensor. The running state recognition unit 13 recognizes the acceleration of the vehicle 20 based on acceleration information from an acceleration sensor. The running state recognition unit 13 recognizes the direction of the vehicle 20 based on yaw rate information from a yaw rate sensor.

[0033] The driver of the vehicle 20 can perform a manual steering operation while the steering control is being executed. The manual steering operation is an operation to override the steering operation while the steering control is being executed. The traveling state recognition unit 13 may determine whether or not a manual steering operation has been performed while the steering control is being executed, for example, based on whether or not a steering operation by the driver has been detected by a steering angle sensor, a steering torque sensor, or the like while the steering control is being executed.

[0034] The route generation unit 14 generates a target route including a target lateral position of the vehicle 20 on the driving lane, based on captured images of the area ahead of the vehicle 20. The target route can be, for example, target lateral position data (lateral position profile) of the vehicle 20 on the driving lane. The target lateral position data may be a set longitudinal position set at predetermined intervals (for example, 1 m) in the traveling direction of the vehicle 20 on the driving lane as the position of the vehicle 20 on the driving lane. The target lateral position is a target position in the width direction of the lane. In this case, the set longitudinal position and the target lateral position may be set together as a single position coordinate. The lateral position profile corresponds to trajectory data represented by associating a target lateral position with each set longitudinal position.

[0035] In addition, when the vehicle 20 is configured to be able to perform automatic steering during automatic driving, the route generation unit 14 may use various methods to generate a path to be used when the vehicle 20 performs automatic driving.

[0036] FIG. 2 is a plan view illustrating the calculation of a learning value for correcting a target lateral position based on the deviation from the target lateral position. The example in FIG. 2 shows a vehicle 20 traveling on a straight road 30 while KLA control is being executed as steering control. The position of vehicle 20 on the traveling lane is center position 21, indicated by a black dot. The target path for KLA control is target path 22, indicated by a dashed straight line. The deviation from the target lateral position corresponds to deviation 23, which is the distance between center position 21 and target path 22 in the lane width direction.

[0037] In the vehicle 20, when no manual steering operation is being performed while steering control is being executed, the actuator 5 is controlled so that the center position 21 is located on the target path 22 of the steering control. In the example of FIG. 2, the driver of the vehicle 20 is performing a manual steering operation while steering control is being executed. For example, this corresponds to a situation in which the target path 22 is offset to the right on the driving lane, and the driver is performing a steering operation to steer the vehicle slightly to the left so that the vehicle 20 travels near the center of the driving lane. The phenomenon in which the target path 22 is offset to the right on the driving lane refers to the effect on the error in the vehicle's lateral position caused by a deviation in the mounting orientation of the camera of the external sensor 2 on the vehicle 20. Note that in FIG. 2, the amount by which the target path 22 is offset to the right on the driving lane is exaggerated.

[0038] When the driver of the vehicle 20 performs a manual steering operation while steering control is being executed, the learning value calculation unit 15 calculates a learning value to correct the target lateral position based on the deviation 23 from the target lateral position that the manual path of the vehicle 20 following the manual steering operation has with respect to the target path 22.

[0039] The manual path of the vehicle 20 according to the manual steering operation corresponds to the movement trajectory of the center position 21 of the vehicle 20 traveling in accordance with the manual steering operation performed during execution of steering control. The manual path of the vehicle 20 extends away from the target path 22 in accordance with the amount of manual steering operation. In the example of FIG. 2, the manual path of the vehicle 20 according to the manual steering operation can be assumed to be, for example, a virtual line along the solid arrow indicating the direction in which the vehicle 20 is traveling.

[0040] The learned value calculation unit 15 calculates a first learned value, which is a learned value when the vehicle 20 is traveling on a straight road, and a second learned value, which is a learned value when the vehicle 20 is traveling on a curved road. The example in FIG. 2 corresponds to a case where the vehicle 20 is traveling on a straight road. Even when the vehicle 20 is traveling on a curved road, the center position 21, the target path 22, and the deviation amount 23 can be determined in the same way as in the example in FIG. 2. The first learned value and the second learned value are learned values ​​for a straight road and a curved road, and can be stored separately in the ECU 10.

[0041] For example, based on a captured image of the area ahead of the vehicle 20, the learning value calculation unit 15 determines whether the vehicle 20 is traveling on a straight road or a curved road in accordance with the shape of a lane marking ahead of the vehicle 20. The learning value calculation unit 15 may also determine whether the vehicle 20 is traveling on a curved road based on the position of the vehicle 20 on a map and map information.

[0042] The learning value calculation unit 15 may calculate a right learning value that is a second learning value when the vehicle 20 is traveling on a right-curving road, and a left learning value that is a second learning value when the vehicle 20 is traveling on a left-curving road. The right learning value and the left learning value are learning values ​​for a right-curving road and a left-curving road, and can be stored separately in the ECU 10.

[0043] For example, when the vehicle 20 is traveling on a curved road, the learning value calculation unit 15 determines whether the vehicle 20 is traveling on a right-curving road in accordance with the curvature direction of the lane markings ahead of the vehicle 20, based on a captured image of the area ahead of the vehicle 20. The learning value calculation unit 15 may determine whether the vehicle 20 is traveling on a right-curving road based on the position of the vehicle 20 on a map and map information.

[0044] The learning value calculation unit 15 may calculate the learning value so that the learning speed of the second learning value is faster than the learning speed of the first learning value. The learning value calculation unit 15 can change the learning speed by changing the calculation cycle of the learning value calculation process.

[0045] For example, when the vehicle 20 is traveling on a straight road, the learning value calculation unit 15 sets the calculation speed of the learning value to a first learning speed by initially setting the calculation period of the learning value calculation process to a first period. For example, when the vehicle 20 is traveling on a curved road, the learning value calculation unit 15 sets the calculation speed of the learning value to a second learning speed by initially setting the calculation period of the learning value calculation process to a second period. The second period is a learning value update period that is shorter than the first period. As a result, the second learning speed becomes a learning value calculation speed that is faster than the first learning speed.

[0046] The vehicle control unit 16 executes steering control based on the target route and the learned value. When a manual steering operation is being performed during the execution of steering control, vehicle control unit 16 may suspend the execution of steering control and wait for the calculation of a learning value by the above-mentioned learning value calculation unit 15. When a manual steering operation is not being performed during the execution of steering control, vehicle control unit 16 may resume the execution of steering control according to the learning value calculated by the above-mentioned learning value calculation unit 15. Vehicle control unit 16 may, for example, correct the position in the vehicle width direction of the target route using the stored learning value, and execute steering control with the corrected position in the vehicle width direction as a target.

[0047] In the example of Fig. 2, it is assumed that a learning value corresponding to deviation amount 23 is stored when a manual steering operation is performed while steering control is being executed. In this case, when manual steering operation is no longer performed while steering control is being executed, the position of target path 22 in the vehicle width direction is corrected by deviation amount 23. As a result, steering control is executed so that center position 21 moves along the solid arrow.

[0048] When the vehicle 20 performs autonomous driving, the vehicle control unit 16 may function to perform autonomous driving of the vehicle 20 based on the route generated by the route generation unit 14. As part of the autonomous driving of the vehicle 20, the above-described steering control may be performed.

[0049] [Example of calculation processing by ECU 10] Next, an example of calculation processing by the ECU 10 will be described. Fig. 3 is a flowchart showing an example of calculation processing of a learning value. The processing shown in Fig. 3 is repeatedly performed at a predetermined cycle, for example, while steering control including steering assistance or automatic steering of the vehicle 20 is being executed.

[0050] 3, in S01, the ECU 10 causes the surrounding environment recognition unit 11 to acquire a captured image of the area ahead of the vehicle 20. The surrounding environment recognition unit 11 uses the camera of the external sensor 2 to acquire a captured image of the area ahead of the vehicle 20.

[0051] In S02, the ECU 10 generates a target route using the route generation unit 14. The route generation unit 14 generates a target route including a target lateral position of the vehicle 20 in the travel lane, based on a captured image of the area ahead of the vehicle 20.

[0052] In S03, the ECU 10 determines, using the learning value calculation unit 15, whether or not the driver is performing a manual steering operation while the steering control is being executed. The learning value calculation unit 15 determines whether or not the driver has performed a manual steering operation while the steering control is being executed, for example, based on whether or not a steering operation by the driver has been detected by a steering sensor or the like of the internal sensor 3 while the steering control is being executed. If the learning value calculation unit 15 determines that the driver has performed a manual steering operation while the steering control is being executed (S03: YES), the ECU 10 interrupts the steering control and proceeds to S04.

[0053] In S04, the ECU 10 calculates the amount of deviation from the target lateral position using the learning value calculation unit 15. The learning value calculation unit 15 calculates the amount of deviation 23, for example, by determining the distance between the center position 21 and the target path 22 in the vehicle width direction.

[0054] In S05, the ECU 10 determines whether or not the vehicle 20 is traveling on a straight road using the learning value calculation unit 15. The learning value calculation unit 15 determines whether or not the vehicle 20 is traveling on a straight road, for example, based on a captured image of the area ahead of the vehicle 20. The learning value calculation unit 15 may determine whether or not the vehicle 20 is traveling on a straight road based on the position of the vehicle 20 on a map and map information.

[0055] If the learning value calculation unit 15 determines that the vehicle 20 is traveling on a straight road (S05: YES), the ECU 10 proceeds to S06. In S06, the ECU 10 causes the learning value calculation unit 15 to calculate a first learning value. The learning value calculation unit 15 calculates the first learning value, which is the learning value when the vehicle 20 is traveling on a straight road, by, for example, multiplying the calculated deviation amount 23 by a predetermined proportionality coefficient and adding the result to the previous value of the first learning value. In S09, the ECU 10 causes the learning value calculation unit 15 to store the learning value. The learning value calculation unit 15 stores the calculated first learning value as the learning value. Thereafter, the ECU 10 ends the processing of FIG. 3.

[0056] On the other hand, if the learning value calculation unit 15 determines that the vehicle 20 is not traveling on a straight road (S05: NO), the ECU 10 proceeds to S07. In S07, the ECU 10 causes the learning value calculation unit 15 to determine whether or not the vehicle 20 is traveling on a curved road. The learning value calculation unit 15 determines whether or not the vehicle 20 is traveling on a curved road, for example, based on a captured image of the area ahead of the vehicle 20. The learning value calculation unit 15 may determine whether or not the vehicle 20 is traveling on a curved road based on the position of the vehicle 20 on a map and map information.

[0057] If the learning value calculation unit 15 determines that the vehicle 20 is traveling on a curved road (S07: YES), the ECU 10 proceeds to S08. In S08, the ECU 10 causes the learning value calculation unit 15 to calculate a second learning value. The learning value calculation unit 15 calculates the second learning value, which is a learning value when the vehicle 20 is traveling on a curved road, for example, by the processing in FIG. 4 described later. In S09, the ECU 10 causes the learning value calculation unit 15 to store the learning value. The learning value calculation unit 15 stores the calculated second learning value as a learning value. Thereafter, the ECU 10 ends the processing in FIG. 3.

[0058] On the other hand, if the learning value calculation unit 15 determines that there is no manual steering operation by the driver while the steering control is being executed (S03: NO), the ECU 10 executes the steering control and ends the processing in Fig. 3. After the learning in Fig. 3 has been executed, the ECU 10 may resume the steering control by the vehicle control unit 16. For example, if the driver who has been manually steering while the steering control is being executed ends the manual steering operation, the vehicle control unit 16 executes the steering control based on the stored learning value and the target route.

[0059] Fig. 4 is a flowchart showing an example of the calculation process of the second learned value in Fig. 3. In S11, the ECU 10 causes the learned value calculation unit 15 to determine whether or not the vehicle 20 is traveling on a road that curves to the right. The learned value calculation unit 15 determines whether or not the vehicle 20 is traveling on a road that curves to the right, for example, based on a captured image of the area ahead of the vehicle 20. The learned value calculation unit 15 may also determine whether or not the vehicle 20 is traveling on a road that curves to the right, based on the position of the vehicle 20 on a map and map information.

[0060] If the learning value calculation unit 15 determines that the vehicle 20 is traveling on a right-curve road (S11: YES), the ECU 10 proceeds to S12. In S12, the ECU 10 causes the learning value calculation unit 15 to calculate a right-curve learned value as the second learned value. For example, the learning value calculation unit 15 multiplies the calculated deviation amount 23 by a predetermined proportional coefficient and adds the result to the previous value of the second learned value to calculate the right-curve learned value, which is the second learned value when the vehicle 20 is traveling on a right-curve road. Thereafter, the ECU 10 ends the processing of FIG. 4 and returns to the processing of FIG. 3.

[0061] On the other hand, if the learning value calculation unit 15 determines that the vehicle 20 is not traveling on a right-hand curve (i.e., the vehicle 20 is traveling on a left-hand curve) (S11: NO), the ECU 10 proceeds to S13. In S13, the ECU 10 causes the learning value calculation unit 15 to calculate a left-hand learned value as the second learned value. For example, the learning value calculation unit 15 multiplies the calculated deviation amount 23 by a predetermined proportional coefficient and adds the result to the previous value of the second learned value to calculate the left-hand learned value, which is the second learned value when the vehicle 20 is traveling on a left-hand curve. Thereafter, the ECU 10 ends the processing of FIG. 4 and returns to the processing of FIG. 3.

[0062] Fig. 5 is a flowchart showing an example of a learning rate setting process. The process shown in Fig. 5 may be repeatedly performed at a predetermined cycle in parallel with the process of Fig. 3 during execution of steering control including steering assistance or automatic steering of the vehicle 20, for example.

[0063] In S21, the ECU 10 determines whether or not the vehicle 20 is traveling on a straight road using the learning value calculation unit 15. The learning value calculation unit 15 determines whether or not the vehicle 20 is traveling on a straight road, for example, based on a captured image of the area ahead of the vehicle 20. The learning value calculation unit 15 may determine whether or not the vehicle 20 is traveling on a straight road based on the position of the vehicle 20 on a map and map information.

[0064] If the learning value calculation unit 15 determines that the vehicle 20 is traveling on a straight road (S21: YES), the ECU 10 proceeds to S22. In S22, the ECU 10 causes the learning value calculation unit 15 to set the calculation speed of the learning value to a first learning speed. For example, the learning value calculation unit 15 sets the calculation speed of the learning value to the first learning speed by setting a predetermined cycle for repeatedly executing the process of FIG. 3 to the first cycle. Thereafter, the ECU 10 ends the process of FIG. 5.

[0065] On the other hand, if the learning value calculation unit 15 determines that the vehicle 20 is not traveling on a straight road (i.e., the vehicle 20 is traveling on a curved road) (S21: NO), the ECU 10 proceeds to S23. In S23, the ECU 10 causes the learning value calculation unit 15 to set the calculation speed of the learning value to a second learning speed. For example, the learning value calculation unit 15 sets the calculation speed of the learning value to the second learning speed by setting a predetermined cycle for repeatedly executing the process of FIG. 3 to a second cycle. Thereafter, the ECU 10 ends the process of FIG. 5.

[0066] According to the steering control device 100 described above, a first learned value for a straight road and a second learned value for a curved road are calculated as learned values ​​for correcting the target lateral position. Here, deviation amount 23, which is an error in the vehicle lateral position on the road 30 on which the vehicle 20 travels, can be caused by a deviation in the mounting posture of the camera that captures the captured image. Deviation amount 23 caused by a deviation in the mounting posture is not necessarily uniform when compared between a straight road and a curved road; for example, it may be larger on a curved road than on a straight road. According to the above configuration, the first learned value for a straight road and the second learned value for a curved road can be calculated as mutually different values. Therefore, a learned value for correcting the target lateral position can be calculated depending on the magnitude of the influence on the error in the vehicle lateral position caused by a deviation in the mounting posture.

[0067] In the steering control device 100, the learning value calculation unit 15 calculates a right learning value, which is the second learning value when the vehicle 20 is traveling on a right-hand curve, and a left learning value, which is the second learning value when the vehicle 20 is traveling on a left-hand curve. As a result, the influence of a deviation in the mounting attitude on the error in the vehicle lateral position is not necessarily uniform when the vehicle 20 is traveling on a right-hand curve and when the vehicle 20 is traveling on a left-hand curve. With the above configuration, the right learning value for a right-hand curve and the left learning value for a left-hand curve can be calculated as different values. Therefore, the second learning values ​​can be calculated corresponding to the difference in the magnitude of the influence of a deviation in the mounting attitude on the error in the vehicle lateral position caused by a deviation on a right-hand curve and a left-hand curve.

[0068] In the steering control device 100, the learning value calculation unit 15 calculates the learning value so that the learning speed of the second learning value is faster than the learning speed of the first learning value. As a result, by making the learning speed faster when traveling on a curved road than when traveling on a straight road, it is possible to correct the vehicle lateral position early on a curved road, which tends to have a large effect on errors in the vehicle lateral position.

[0069] [Variations] Although the embodiments of the present invention have been described above, the present invention is not limited to the above-described embodiments. The present invention can be embodied in various forms, including the above-described embodiments, with various modifications and improvements made based on the knowledge of those skilled in the art.

[0070] In the above embodiment, the learning value calculation unit 15 calculates a right learning value that is the second learning value when the vehicle 20 is traveling on a right-curving road and a left learning value that is the second learning value when the vehicle 20 is traveling on a left-curving road, but this example is not essential. The second learning value may be a common learning value for right-curving roads and left-curving roads. The learning value calculation unit 15 only needs to separately calculate at least a first learning value that is the learning value when the vehicle 20 is traveling on a straight road and a second learning value that is the learning value when the vehicle 20 is traveling on a curved road.

[0071] In the above embodiment, the learning value calculation unit 15 calculates the learning value so that the learning speed of the second learning value is faster than the learning speed of the first learning value, but this example is not essential. The learning value calculation unit 15 may calculate the first learning value and the second learning value at the same learning speed.

[0072] In the above embodiment, the vehicle 20 is configured to be able to perform autonomous driving, but this example is not essential. The steering control device 100 may be configured to be able to perform steering control including steering assist or automatic steering in a vehicle 20 that is configured not to perform autonomous driving. In this case, components for performing autonomous driving that are not used for steering control including steering assist or automatic steering (such as the GPS receiver 1 and the map database 4) may be omitted. [Explanation of symbols]

[0073] 14... Path generation unit, 15... learning value calculation unit, 16... vehicle control unit (steering control unit), 20... vehicle, 22... target path, 23... deviation amount, 100... steering control device.

Claims

1. A steering control device that performs steering control including steering assistance or automatic steering of a vehicle, a route generation unit that generates a target route including a target lateral position of the vehicle in a travel lane based on a captured image of a front of the vehicle; a learning value calculation unit that calculates, when a driver of the vehicle is performing a manual steering operation during execution of the steering control, a learning value to correct the target lateral position based on a deviation amount from the target lateral position that a manual path of the vehicle in accordance with the manual steering operation has with respect to the target path; a steering control unit that executes the steering control based on the target route and the learned value, The learning value calculation unit calculates a first learning value that is the learning value when the vehicle is traveling on a straight road, and a second learning value that is the learning value when the vehicle is traveling on a curved road.

2. 2. The steering control device according to claim 1, wherein the learning value calculation unit calculates a right learning value that is the second learning value when the vehicle is traveling on a right-curving road, and a left learning value that is the second learning value when the vehicle is traveling on a left-curving road.

3. The steering control device according to claim 1 or 2, wherein the learning value calculation unit calculates the learning value such that a learning speed of the second learning value is faster than a learning speed of the first learning value.

Images