Vehicle control device and vehicle control method

Abstract

In one aspect of a vehicle control device and a vehicle control method according to the present invention: information relating to first left and right boundary lines indicating a range over which a vehicle can travel is acquired; the first left and right boundary lines are subjected to smoothing processing to obtain information relating to second left and right boundary lines after the smoothing processing; the information relating to the second left and right boundary lines is stored, and processing for obtaining a travel route for the vehicle is executed using the information relating to the second left and right boundary lines; and as the smoothing processing, the information relating to the first left and right boundary lines and the information relating to past second left and right boundary lines that has been stored are each weighted and combined. This configuration makes it is possible to fit the left and right boundary lines even on a road having a complex shape, while suppressing the calculation load.

Description

Vehicle control device and vehicle control method

[0001] The present invention relates to a vehicle control device and a vehicle control method.

[0002] The vehicle control device of Patent Document 1 extracts roadway defining objects such as lane marks, guardrails, and curbs that define the drivable range of the roadway based on image information from a camera, and generates left and right recognition lines by polynomial approximating these roadway defining objects.

[0003] Japanese Patent Application Laid-Open No. 2018-062244

[0004] However, when lane marks and the like are extracted from external recognition information and the extracted lane marks and the like are used to obtain the left and right boundary lines by polynomial approximation, it becomes difficult to approximate the left and right boundary lines on roads with large curvatures or complex shapes, and there is a problem in that the amount of calculation required for the process of obtaining the left and right boundary lines becomes large.

[0005] The present invention has been made in consideration of the current situation, and its purpose is to provide a vehicle control device and a vehicle control method that can fit left and right boundary lines even on roads with complex shapes while reducing the calculation load.

[0006] In one aspect, the vehicle control device of the present invention is a vehicle control device mounted on a vehicle, and includes a boundary line information acquisition unit that acquires information on a first left and right boundary line that indicates the drivable range of the driving road on which the vehicle is driving, a fluctuation suppression processing unit that performs a smoothing process on the first left and right boundary line to obtain information on the second left and right boundary line after the smoothing process, a memory unit that stores information on the second left and right boundary line, and a driving route generation unit that uses the information on the second left and right boundary line to obtain a driving route on which the vehicle will drive, and the fluctuation suppression processing unit performs the smoothing process by weighting and combining the information on the first left and right boundary line and past information on the second left and right boundary line stored in the memory unit.

[0007] In one aspect, the vehicle control method of the present invention is a vehicle control method executed by a control unit mounted on a vehicle, and includes the steps of acquiring information on a first left-right boundary line indicating the drivable range of the road on which the vehicle is traveling, performing a smoothing process on the first left-right boundary line to obtain information on a second left-right boundary line after the smoothing process, storing the information on the second left-right boundary line, and using the information on the second left-right boundary line to obtain the driving route on which the vehicle is traveling, and the step of obtaining the information on the second left-right boundary line performs the smoothing process by weighting and combining the information on the first left-right boundary line and stored past information on the second left-right boundary line.

[0008] According to the above invention, it is possible to fit left and right boundary lines even for roads with complex shapes while suppressing the calculation load.

[0009] 1 is a block diagram showing a vehicle control system. FIG. 1 is a block diagram showing a first embodiment of a process for smoothing left and right boundary lines. FIG. 2 is a flowchart showing the flow of a smoothing program of the first embodiment. FIG. 3 is a diagram showing a weighting calculation in the smoothing process. FIG. 4 is a diagram showing the correlation between the distance from the vehicle and a weighting coefficient α. FIG. 5 is a diagram showing how the left and right boundary lines fluctuate. FIG. 6 is a diagram showing how the left and right boundary lines are corrected after the smoothing process. FIG. 7 is a diagram showing the correlation between the vehicle speed and a correction value α2. FIG. 8 is a diagram showing the correlation between the road curvature and a correction value α3. FIG. 9 is a diagram showing the correlation between the vehicle vibration and a correction value α4. FIG. 10 is a diagram showing the correlation between the cycle of the smoothing process and a correction value α5. FIG. 11 is a diagram showing the correlation between the accuracy of recognition of the left and right boundary lines and a correction value α6. FIG. 12 is a diagram showing another aspect of the correlation between the distance from the vehicle and the weighting coefficient α. FIG. 13 is a flowchart showing the flow of a program for calculating the weighting coefficient α. FIG. 14 is a diagram illustrating the correlation between the distance from the vehicle and the correction value α2. FIG. 15 is a diagram illustrating the correlation between the distance from the vehicle and the correction value α3. FIG. 16 is a diagram illustrating the correlation between the distance from the vehicle and the correction value α4. 1 is a diagram illustrating the correlation between the distance from the vehicle and the correction value α5. FIG. 2 is a diagram illustrating the correlation between the distance from the vehicle and the correction value α6. FIG. 3 is a diagram illustrating the change characteristic of the weighting coefficient α with respect to the distance from the vehicle. FIG. 4 is a block diagram showing a first embodiment of a smoothing process for left and right boundary lines. FIG. 5 is a flowchart showing the flow of a smoothing program of a second embodiment. FIG. 6 is a diagram illustrating a threshold value (safety margin) used in deviation determination. FIG. 7 is a diagram illustrating the correlation between the deviation amount and a threshold value. FIG. 8 is a diagram illustrating the characteristic of changing the deviation amount threshold value according to the distance from the vehicle. FIG. 9 is a diagram illustrating the determination of the ratio of the deviation range to the entire boundary line. FIG. 10 is a diagram illustrating an example in which the ratio of the deviation range to the entire boundary line is different.

[0010] DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS A vehicle control system 200 including a vehicle control device according to an embodiment of the present invention will now be described with reference to the accompanying drawings.

[0011] The vehicle control system 200 is a system mounted on a vehicle 100 such as a four-wheeled automobile to control the motion of the vehicle 100, and includes an external environment recognition unit 300, an automatic driving control unit 500, a vehicle motion control unit 600, and an actuator unit 700. The automatic driving control unit 500 and the vehicle motion control unit 600 constitute a vehicle control device, and the automatic driving control unit 500 and the vehicle motion control unit 600 execute a vehicle control method.

[0012] The external environment recognition unit 300 includes a GPS (Global Positioning System) receiving unit 310, a map database 320, a road-to-vehicle communication device 330, a camera 340, a radar 350, and a LiDAR (Light Detection and Ranging, Laser Imaging Detection and Ranging) 360. The external environment recognition unit 300 acquires external environment information of the vehicle 100 (in other words, information about the driving environment of the vehicle 100) from each device.

[0013] The GPS receiver 310 receives signals from GPS satellites to measure the latitude and longitude of the location of the vehicle 100. The map database 320 is a collection of map information including road locations, road shapes, intersection locations, etc., and is created in a storage device installed in the vehicle 100.

[0014] The road-to-vehicle communication device 330 transmits information about the vehicle 100 to a roadside device and receives road traffic information such as curves and intersections from the roadside device. The external environment recognition unit 300 may include a vehicle-to-vehicle communication device that communicates between the vehicle 100 and other vehicles and acquires road traffic information, behavior information of other vehicles, and the like from the other vehicles with which the vehicle 100 is communicating.

[0015] The camera 340 is a stereo camera, a monocular camera, a 360° camera, or the like, and captures images of the surroundings of the vehicle 100 to acquire image information of the surroundings of the vehicle 100. The radar 350 and the LiDAR 360 detect objects around the vehicle 100 and output information about the detected objects.

[0016] The autonomous driving control unit 500 is an electronic control device mainly including a microcomputer 540 as a control section that performs calculations based on input information and outputs the calculation results. The microcomputer 540 includes an MPU (Microprocessor Unit), ROM (Read Only Memory), RAM (Random Access Memory), etc., which are not shown. The microcomputer 540 of the autonomous driving control unit 500 acquires external environment information recognized by the external environment recognition section 300, and outputs target commands calculated based on the acquired various information to the vehicle motion control unit 600.

[0017] The microcomputer 540 of the autonomous driving control unit 500 includes, as software, the functional units of a surrounding situation recognition unit 510, an action planning unit 520, and a goal generation unit 530. The surrounding situation recognition unit 510 recognizes the surrounding situation of the vehicle 100 based on an external environment recognition signal from the external environment recognition unit 300, etc.

[0018] The surrounding conditions of the vehicle 100 recognized by the surrounding conditions recognition unit 510 include information such as the curvature of the road, the road surface cant, the road surface gradient, the friction coefficient μ of the road surface, the positions of left and right lane markers (road markings), the positions of left and right road edges, moving objects, and stationary objects. Here, the information on the positions of the left and right lane markers and the positions of the left and right road edges, as well as information on moving objects and stationary objects, is information on the left and right boundary lines that indicate the drivable range of the road on which the vehicle 100 is traveling.

[0019] That is, the surrounding situation recognition unit 510 includes a boundary line information acquisition unit 510a that acquires information about left and right boundary lines that indicate the driving range of the road on which the vehicle 100 is traveling. Note that the moving objects recognized by the surrounding situation recognition unit 510 include pedestrians, bicycles, motorcycles, other vehicles, etc., and the stationary objects include objects that have fallen on the road, traffic signals, guardrails, curbs, road signs, trees, billboards, etc.

[0020] The action planning unit 520 acquires the recognition results from the surrounding situation recognition unit 510 and creates an action plan for the vehicle 100, including the selection of a driving lane and the selection of a traveling direction at intersections and branching points. Then, the goal generation unit 530 determines a target command to be output to the vehicle motion control unit 600 based on the surrounding situation of the vehicle 100 recognized and determined by the surrounding situation recognition unit 510 and the action plan for the vehicle 100 created by the action planning unit 520.

[0021] Here, the target command that the target generation unit 530 issues to the vehicle motion control unit 600 includes a command specifying the driving area ahead of the vehicle 100. Note that, as the driving area is defined by left and right boundary lines, as will be described later, the command for the driving area is also a command for the left and right boundary lines. In this way, the microcomputer 540 of the autonomous driving control unit 500 recognizes and judges the surrounding situation of the vehicle 100 based on the external environment information of the vehicle 100 acquired by the external environment recognition unit 300, and outputs instruction information for the driving area (left and right boundary lines) as a target command.

[0022] Vehicle motion control unit 600 is an electronic control device mainly including a microcomputer 630 as a control section that performs calculations based on input information and outputs the calculation results, and microcomputer 630 includes an MPU, ROM, RAM, etc. (not shown). Microcomputer 630 of vehicle motion control unit 600 obtains target commands including instruction information for the driving area (left and right boundary lines) from automatic driving control unit 500, and outputs control commands to actuator section 700 for causing vehicle 100 to drive in accordance with the target commands.

[0023] The microcomputer 630 of the vehicle motion control unit 600 includes, as software, the functional units of a driving route generation unit 610 and a following control unit 620. The driving route generation unit 610 plans a driving route for the vehicle 100 within a driving area instructed by the autonomous driving control unit 500, and also plans the speed at which the vehicle 100 will travel along the driving route.

[0024] For example, the driving route generation unit 610 effectively utilizes the width of the driving area (drivable range) instructed by the autonomous driving control unit 500, and by making the curvature of the route as gentle as possible on turning roads, plans a driving route that can suppress the acceleration and jerk of the vehicle 100. Furthermore, the driving route generation unit 610 plans a speed so that the acceleration that occurs when the vehicle 100 travels along the planned driving route changes small and gradually.

[0025] The following control unit 620 acquires information on the traveling longitude and speed planned by the traveling route generation unit 610. Then, the following control unit 620 calculates control commands, specifically steering commands (steering angle command, steering force command) and acceleration / deceleration commands (driving force command, braking force command), for causing the vehicle 100 to follow the traveling route and speed set by the traveling route generation unit 610, and outputs the calculated control commands to the actuator unit 700.

[0026] Actuator section 700 includes an internal combustion engine 710 and a motor inverter 720 that generate driving force for vehicle 100, a braking device 730 that applies braking force to vehicle 100, an electronically controlled power steering device 740 that changes the direction of travel of vehicle 100, and an electronically controlled suspension 750 that can adjust damping force and vehicle height. Actuator section 700 generates driving force, braking force, steering force, etc. in response to control commands from tracking control section 620.

[0027] Here, the target generator 530 fits the variation in the left and right boundary lines by smoothing the left and right boundary lines that define the driveable area of ​​the vehicle 100, which are acquired by the boundary line information acquirer 510a. Then, the target generator 530 outputs the information on the left and right boundary lines after the smoothing process to the drive path generator 610 as a command for the drive area.

[0028] The left and right boundary lines are not limited to a single lane marking, but may be markings on both the left and right ends that include multiple lanes, taking into consideration lane changes by the vehicle 100. Furthermore, when there are no markings (lane markers) or when they are unclear, the left and right boundary lines may be lines that indicate the edge of the road, such as a step or a curb.

[0029] Furthermore, even if there are dividing lines, if there is an area within the lane where travel is prohibited due to an obstacle such as a parked vehicle, the left and right boundary lines serve as boundary lines that define a travel area that avoids this obstacle. In this way, the left and right boundary lines are boundary lines that indicate the range within which vehicle 100 can actually travel while ensuring safety depending on the situation.

[0030] 2 is a block diagram showing a first embodiment of the process for smoothing left and right boundary lines. The target generator 530 includes a fluctuation suppression processor 530a that smooths information about left and right boundary lines (hereinafter referred to as first left and right boundary lines LL1, LR1) acquired by the boundary line information acquisition unit 510a, and a storage unit 530b that stores information about left and right boundary lines (hereinafter referred to as second left and right boundary lines LL2, LR2) that have been smoothed by the fluctuation suppression processor 530a. The storage unit 530b saves the information about the second left and right boundary lines LL2, LR2 in a memory such as a RAM.

[0031] The fluctuation suppression processing unit 530a performs a smoothing process by weighting and combining information on the first left and right boundary lines LL1, LR1 (left and right boundary lines before the smoothing process) acquired by the boundary line information acquisition unit 510a and information on the past second left and right boundary lines LL2p, LR2p (left and right boundary lines after the past smoothing process) stored in the storage unit 530b, thereby acquiring information on the second left and right boundary lines LL2, LR2 after the smoothing process, in which fluctuations due to noise and the like have been suppressed. The fluctuation suppression processing unit 530a outputs the information on the second left and right boundary lines LL2, LR2 after the smoothing process, in other words, information on the driving range, to the driving route generation unit 610. The driving route generation unit 610 then generates a driving route within the specified driving range based on the information on the second left and right boundary lines LL2, LR2.

[0032] 3 is a flowchart showing the flow of a left / right boundary line smoothing program executed by the microcomputer 540 of the autonomous driving control unit 500. In step S11 (boundary line information acquisition unit 510a), the microcomputer 540 acquires information about the first left / right boundary lines LL1, LR1 based on the external environment recognition result by the external environment recognition unit 300.

[0033] Next, in step S12 (fluctuation suppression processing unit 530a), the microcomputer 540 performs a smoothing process by weighting and combining the acquired information on the first left-right boundary lines LL1, LR1 and the past information on the second left-right boundary lines LL2p, LR2p, and acquires information on the second left-right boundary lines LL2, LR2 in which fluctuations due to noise, etc. have been suppressed. Then, in step S13 (travel path generation unit 610, storage unit 530b), the microcomputer 540 generates a travel path based on the information on the second left-right boundary lines LL2, LR2 after the smoothing process, and also stores the currently acquired information on the second left-right boundary lines LL2, LR2 in memory.

[0034] 4 is a diagram showing details of the smoothing process in the fluctuation suppression processing unit 530a. Note that the information on the left and right boundary lines is expressed as a coordinate sequence of points arranged in order from the vicinity of the vehicle 100 toward the front of the vehicle 100.

[0035] The fluctuation suppression processing unit 530a acquires the previous values ​​LL2p(n-1), LR2p(n-1) of the past second left-right boundary lines LL2p, LR2p stored in the storage unit 530b and the information on the first left-right boundary lines LL1, LR1 acquired by the boundary line information acquisition unit 510a.The fluctuation suppression processing unit 530a then determines the points (asterisks in FIG. 4) on the first left-right boundary lines LL1, LR1 that are closest to each point (black circles in FIG. 4) of the second left-right boundary lines LL2p(n-1), LR2p(n-1).

[0036] Next, the fluctuation suppression processing unit 530a combines each point (black dots in FIG. 4) of the second left-right boundary lines LL2p(n-1), LR2p(n-1) with the nearest point (asterisk symbol in FIG. 4) on the first left-right boundary lines LL1, LR1 using a weighting coefficient α (0≦α≦1) to obtain a point (point marked with an x ​​in FIG. 4) that defines the second left-right boundary lines LL2, LR2 after the smoothing process. Point after smoothing process = (1−α) × previous value of point after smoothing process + α × nearest point Then, the fluctuation suppression processing unit 530a outputs information about the second left-right boundary lines LL2, LR2 after the smoothing process to the travel path generation unit 610 and the memory unit 530b.

[0037] Since the smoothing process described above is performed on each control point, it is not affected by the shape of the left and right boundary lines (in other words, the way the control points are connected), and can accommodate complex boundary line shapes. On the other hand, in the case of fitting using polynomial approximation, in order to accommodate complex boundary line shapes, the degree of the polynomial must be increased, which increases the amount of calculation. In other words, the smoothing process described above makes it possible to fit left and right boundary lines even on roads with complex shapes while reducing the calculation load on the microcomputer 540.

[0038] Furthermore, when a driving route is generated using an optimization algorithm to minimize an evaluation function that includes ride comfort within the drivable area between the left and right boundaries, fluctuations in the drivable area are directly linked to fluctuations in the driving route. Therefore, if the setting of the drivable area can be stabilized by smoothing processing, fluctuations in the driving route can be suppressed and ride comfort can be improved. Furthermore, because the optimization algorithm requires large-scale calculations, if the smoothing processing reduces the calculation load on the microcomputer, computing resources can be allocated to the calculations of the optimization algorithm, making it possible to generate an optimized route with a more comfortable ride.

[0039] Here, the fluctuation suppression processing unit 530a can set the weighting coefficient α (weight) to a fixed value, but by variably setting the weighting coefficient α in accordance with various conditions in the smoothing process, the fitting accuracy of the left and right boundary lines can be improved. For example, the fluctuation suppression processing unit 530a can vary the weighting coefficient α in accordance with the distance from the vehicle 100 to the point on the left and right boundary lines to be smoothed.

[0040] 5 is a diagram illustrating an example of the correlation between the distance from the vehicle 100 and the weighting coefficient α. Here, the fluctuation suppression processing unit 530a sets the weighting coefficient α to 0 near the vehicle 100, and increases the weighting coefficient α as the distance from the vehicle 100 increases (as the distance from the vehicle 100 increases), bringing the weighting coefficient α closer to the upper limit of 1.0. In other words, as the distance from the vehicle 100 increases, the fluctuation suppression processing unit 530a increases the weighting assigned to the first left and right boundary lines LL1, LR1 acquired by the boundary line information acquisition unit 510a, and decreases the weighting assigned to the past second left and right boundary lines LL2p, LR2p relatively.

[0041] The information about the first left-right boundary lines LL1, LR1 acquired by the boundary line information acquisition unit 510a is superimposed with fluctuations due to sensor vibrations, recognition errors, etc. The fluctuations superimposed on the information about the first left-right boundary lines LL1, LR1 are generally smaller closer to the vehicle 100 and larger the farther away from the vehicle 100. For this reason, the reliability of the information about the first left-right boundary lines LL1, LR1 acquired by the boundary line information acquisition unit 510a tends to be higher closer to the vehicle 100 and lower the farther away from the vehicle 100.

[0042] Therefore, it is desirable to decrease the weighting coefficient α near the vehicle 100 to increase the weighting (1-α) of the past second left-right boundary lines LL2p, LR2p (more specifically, the previous values ​​LL2p(n-1), LR2p(n-1)) and thereby stabilize the second left-right boundary lines LL2, LR2 after the smoothing process as much as possible. On the other hand, it is desirable to increase the weighting coefficient α farther from the vehicle 100, thereby increasing the weighting of the first left-right boundary lines LL1, LR1 acquired by the boundary line information acquisition unit 510a and thereby increasing the amount of correction, thereby quickly converging the left-right boundary lines after the smoothing process to their true values. Therefore, as shown in FIG. 5 , the fluctuation suppression processing unit 530a increases the weighting coefficient α as the distance from the vehicle 100 increases, thereby achieving both stability of the second left-right boundary lines LL2, LR2 and convergence of the second left-right boundary lines LL2, LR2 to their true values.

[0043] 6 shows how the amount of fluctuation (in other words, the error) of the first left-right boundary lines LL1, LR1 caused by the superposition of random noise increases as the distance from the vehicle 100 increases. Note that, because FIG. 6 assumes completely random noise, the first left-right boundary lines LL1, LR1 converge, on average, to the vicinity of the actual left-right boundary lines.

[0044] 7 illustrates an example of changes in the correction amount when the weighting coefficient α is varied according to the distance from the vehicle 100, specifically, when the weighting of the first left-right boundary lines LL1, LR1 acquired by the boundary line information acquisition unit 510a is increased as the distance from the vehicle 100 increases, thereby increasing the correction amount of the smoothing process. Points on the first left-right boundary lines LL1, LR1 are corrected by the smoothing process with a larger correction amount the farther the point is from the vehicle 100, and are corrected multiple times as the point approaches the vehicle 100. As a result, the second left-right boundary lines LL2, LR2 obtained by the smoothing process converge to the vicinity of the actual left-right boundary lines near the vehicle 100 and stabilize at values ​​close to the actual left-right boundary lines (true values).

[0045] The condition for varying the weighting coefficient α is not limited to the distance from the vehicle 100, and the fluctuation suppression processing unit 530a can vary the weighting coefficient α by combining a plurality of conditions. Specifically, the fluctuation suppression processing unit 530a can vary the weighting coefficient α based on at least one of the distance from the vehicle 100, the speed (vehicle speed) of the vehicle 100, the curvature of the road, the vibration of the vehicle 100, the cycle of the smoothing processing, and the accuracy of recognition of the first left and right boundary lines LL1 and LR1.

[0046] Hereinafter, as one mode of processing for varying the weighting coefficient α by combining multiple conditions, a process for varying the weighting coefficient α based on all of the distance from the vehicle 100, the speed (vehicle speed) of the vehicle 100, the curvature of the road, the vibration of the vehicle 100, the cycle of the smoothing process, and the accuracy of recognition of the first left and right boundary lines LL1 and LR1 will be described. Here, the fluctuation suppression processing unit 530a sets the weighting coefficient α based on the distance from the vehicle 100 to a basic value α1 (0≦α1≦1.0) of the weighting coefficient α.

[0047] The fluctuation suppression processing unit 530a also sets correction values ​​α2, α3, α4, α5, and α6 for the basic value α1 based on other conditions, namely, the speed of the vehicle 100, the curvature of the road, the vibration of the vehicle 100, the cycle of the smoothing processing, and the accuracy of recognition of the first left and right boundary lines LL1 and LR1. The fluctuation suppression processing unit 530a then calculates the final weighting coefficient α (0≦α≦1.0) as α=α1×α2×α3×α4×α5×α6.

[0048] As will be described in detail later, the correction values ​​α2, α3, α4, α5, and α6 are variably set within a range of 0 to 2.0, and when the correction values ​​α2, α3, α4, α5, and α6 exceed 1.0, the weighting coefficient α1, which is the base value, is increased, whereas when the correction values ​​α2, α3, α4, α5, and α6 are below 1.0, the weighting coefficient α1, which is the base value, is decreased. Furthermore, when the weighting coefficient α calculated from the weighting coefficient α1 and the correction values ​​α2, α3, α4, α5, and α6 exceeds 1.0, the fluctuation suppression processing unit 530a resets the weighting coefficient α to 1.0, thereby limiting the weighting coefficient α so that it does not exceed the upper limit of 1.0.

[0049] Next, the characteristics of the weighting coefficient α1 and the correction values ​​α2, α3, α4, α5, and α6 will be described. The basic value α1 (0≦α1≦1.0) of the weighting coefficient α based on the distance from the vehicle 100 is set to 0 in the vicinity of the vehicle 100, as shown in FIG. 5, and increases toward 1.0 as the distance from the vehicle 100 increases.

[0050] That is, the fluctuation suppression processing unit 530a increases the weighting on the first left-right boundary lines LL1, LR1 at positions far from the vehicle 100 where large fluctuation amounts overlap, thereby increasing the amount of correction in the smoothing processing and ensuring the convergence of the second left-right boundary lines LL2, LR2 to their true values ​​after the smoothing processing. On the other hand, near the vehicle 100, the fluctuation suppression processing unit 530a increases the weighting on the second left-right boundary lines LL2p, LR2p obtained in the past by the smoothing processing, thereby improving the stability of the second left-right boundary lines LL2, LR2.

[0051] 8 is a diagram illustrating an example of the correlation between the speed (vehicle speed) [m / s] of the vehicle 100 and the correction value α2 (0≦α2≦2.0). The correction value α2 is set to 0 when the vehicle 100 is stopped and the vehicle speed is 0 [m / s], and is set to a larger value as the vehicle speed increases.

[0052] This is because the higher the vehicle speed, the fewer times the smoothing process is performed per certain distance, which delays the convergence of the left and right boundary lines to their true values ​​after the smoothing process.The fluctuation suppression processing unit 530a increases the weighting assigned to the first left and right boundary lines LL1, LR1 acquired by the boundary line information acquisition unit 510a as the vehicle speed increases, thereby increasing the amount of correction in the smoothing process and ensuring the convergence of the second left and right boundary lines LL2, LR2 to their true values ​​after the smoothing process.

[0053] 9 is a diagram illustrating the correlation between the curvature [1 / m] of the road on which the vehicle 100 is traveling and the correction value α3 (0≦α3≦2.0). The road curvature is acquired from map information and forward recognition results obtained by a camera. The correction value α3 is set to 0 for straight roads, and is set to a larger value as the road curvature (in other words, the curvature radius) becomes larger and the curve becomes sharper.

[0054] This is because the sharper the curve, the greater the fluctuation of the first left and right boundary lines LL1, LR1. The fluctuation suppression processing unit 530a increases the weighting of the first left and right boundary lines LL1, LR1 acquired by the boundary line information acquisition unit 510a as the curvature of the road increases, thereby increasing the amount of correction in the smoothing process, thereby ensuring the convergence of the second left and right boundary lines LL2, LR2 to their true values ​​after the smoothing process.

[0055] 10 is a diagram illustrating the correlation between the vibration of the vehicle 100 and the correction value α4 (0≦α4≦2.0). The vibration of the vehicle 100 includes vibration of displacement in the front-rear, left-right, and up-down directions, as well as vibration of the rotation angle around three axes, and is detected by an acceleration sensor and an angular velocity sensor. The correction value α4 is set to 0 when there is no vibration of the vehicle 100, and is set to a larger value as the vibration of the vehicle 100 increases.

[0056] This is because the external environment recognition unit 300, such as the camera 340, also vibrates in conjunction with the vibration of the vehicle 100, and the greater the vibration of the vehicle 100 (external environment recognition unit 300), the greater the amount of fluctuation in the first left and right boundary lines LL1, LR1 obtained using the external environment recognition unit 300. Then, the greater the vibration of the vehicle 100, the greater the fluctuation suppression processing unit 530a assigns to the first left and right boundary lines LL1, LR1 acquired by the boundary line information acquisition unit 510a, and increases the correction amount of the smoothing processing, thereby ensuring convergence of the second left and right boundary lines LL2, LR2 to their true values ​​after the smoothing processing.

[0057] 11 is a diagram illustrating an example of the correlation between the calculation period of the smoothing process (in other words, the update period of the second left and right boundary lines LL2, LR2) [ms] and the correction value α5 (0≦α5≦2.0). If the calculation period of the smoothing process is constant, the fluctuation suppression processing unit 530a will provide the correction value α5 as a fixed value. However, if the calculation period of the smoothing process is variable depending on conditions such as vehicle speed and road shape, the fluctuation suppression processing unit 530a will obtain information about the calculation period at that time and set the correction value α5.

[0058] The correction value α5 is set to a larger value as the calculation cycle of the smoothing process becomes longer (in other words, as the calculation frequency, which is the reciprocal of the calculation cycle, becomes lower). This is because the longer the calculation cycle of the smoothing process, the fewer times the smoothing process is performed per certain time period, which delays the convergence of the second left and right boundary lines LL2, LR2 to their true values. The fluctuation suppression processing unit 530a then increases the weighting of the first left and right boundary lines LL1, LR1 acquired by the boundary line information acquisition unit 510a as the calculation cycle of the smoothing process becomes longer, thereby ensuring the convergence of the second left and right boundary lines LL2, LR2 to their true values ​​after the smoothing process.

[0059] 12 is a diagram illustrating an example of the correlation between the accuracy (in other words, reliability) of the recognition of the first left-right boundary lines LL1 and LR1 and the correction value α6 (0≦α6≦2.0). The accuracy of the recognition of the first left-right boundary lines LL1 and LR1 is evaluated based on the performance of the camera 340 used to recognize the first left-right boundary lines LL1 and LR1, the degree of blurring of the lines painted on the road to indicate the left-right boundary lines, the brightness around the vehicle 100, weather conditions, etc.

[0060] The correction value α6 is set to a larger value as the boundary information acquisition unit 510a recognizes the first left-right boundary lines LL1 and LR1 with a lower degree of accuracy. Conversely, the correction value α6 is set to a smaller value as the boundary information acquisition unit 510a recognizes the first left-right boundary lines LL1 and LR1 with a higher degree of accuracy. This is because if the weighting of the second left-right boundary lines LL2p and LR2p after the previous smoothing process is increased when the recognition accuracy of the first left-right boundary lines LL1 and LR1 is low, the error (variation amount) will not be corrected easily, and driving control will continue based on the second left-right boundary lines LL2p and LR2p with a large error. The fluctuation suppression processing unit 530a increases the weighting of the first left-right boundary lines LL1 and LR1 as the recognition accuracy of the first left-right boundary lines LL1 and LR1 decreases, thereby increasing the correction amount and ensuring convergence of the second left-right boundary lines LL2 and LR2 to their true values ​​after the smoothing process.

[0061] The change characteristics of the weighting coefficient α1 and the correction values ​​α2, α3, α4, α5, and α6 are not limited to characteristics that change continuously with changes in the distance from the vehicle 100, the vehicle speed, etc. Fig. 13 illustrates an example of a characteristic in which the weighting coefficient α1 changes discretely in a stepwise manner with changes in the speed of the vehicle 100, and the correction values ​​α2, α3, α4, α5, and α6 can also have characteristics that change discretely in a stepwise manner.

[0062] The flowchart in Figure 14 shows the flow of a program for calculating the weighting coefficient α executed by the microcomputer 540 (fluctuation suppression processing unit 530a). In step S21, the microcomputer 540 selects one conversion table to refer to from the conversion table list used to calculate the weighting coefficient α, that is, from the conversion tables shown in Figures 5 and 8-12. For example, the microcomputer 540 sequentially selects conversion tables so as to set the weighting coefficient α1, and the correction values ​​α2, α3, α4, α5, and α6 in this order.

[0063] Next, in step S22, the microcomputer 540 acquires the source information in the currently referenced conversion table, i.e., any one of the distance from the vehicle 100, the speed of the vehicle 100, the curvature of the road, the vibration of the vehicle 100, the cycle of the smoothing process, and the accuracy of recognition of the first left and right boundary lines LL1, LR1. Then, in step S23, the microcomputer 540 converts the source information acquired in step S22 into the weighting coefficient α1 or the correction values ​​α2, α3, α4, α5, α6 using the referenced conversion table.

[0064] In the calculation process of the correction values ​​α2, α3, α4, α5, and α6 in step S23, the microcomputer 540 calculates the correction values ​​α2, α3, α4, α5, and α6 for each distance from the vehicle 100. Figures 15 to 19 show examples of the results of calculating the correction values ​​α2, α3, α4, α5, and α6 for each distance from the vehicle 100.

[0065] Here, the speed of the vehicle 100, the vibration of the vehicle 100, and the period of the smoothing process are values ​​that are determined regardless of the distance from the vehicle 100. For this reason, the microcomputer 540 sets the correction values ​​α2, α4, and α5 based on the speed of the vehicle 100, the vibration of the vehicle 100, and the period of the smoothing process to constant values ​​found from the conversion tables in Figures 8, 10, and 11, regardless of the distance from the vehicle 100, as shown in Figures 15, 17, and 18.

[0066] On the other hand, the curvature of the road changes depending on the distance from the vehicle 100. Therefore, the microcomputer 540 determines the correction value α3 by referring to the conversion table of Fig. 9 based on the road curvature for each distance from the vehicle 100, and sets a conversion table that determines the correction value α3 for each distance from the vehicle 100, as shown in Fig. 16.

[0067] Furthermore, because the degree of blurring of lane markers (road markings) varies depending on the location, the accuracy of recognition of the first left and right boundary lines LL1, LR1 may also vary depending on the distance from the vehicle 100. For this reason, the microcomputer 540 determines the correction value α6 by referring to the conversion table of Fig. 12 based on the accuracy for each distance from the vehicle 100, and sets a conversion table that determines the correction value α6 for each distance from the vehicle 100, as shown in Fig. 19.

[0068] Next, in step S24, the microcomputer 540 calculates the previous value α of the weighting coefficient α. p The weighting coefficient α is updated by multiplying the weighting coefficient α1 calculated based on the distance from the vehicle 100 this time or by one of the correction values ​​α2, α3, α4, α5, and α6. The initial value of the weighting coefficient α is set to 1.0.

[0069] Then, in step S25, the microcomputer 540 determines whether or not the selection of all tables has been completed. In other words, in step S25, the microcomputer 540 determines whether or not the weighting coefficient α has been calculated based on the weighting coefficient α1 and all of the correction values ​​α2, α3, α4, α5, and α6.

[0070] If the selection of all conversion tables has not been completed, the microcomputer 540 returns from step S25 to step S21 to select the next conversion table to be referenced. That is, when the microcomputer 540 sequentially selects conversion tables in the order of α1, α2, α3, α4, α5, and α6, for example, the microcomputer 540 sets α=α in step S24. p ×α1=1.0×α1, α=α p ×α2=(1.0×α1)×α2, α=α p ×α3=(1.0×α1×α2)×α3, α=α p ×α4=(1.0×α1×α2×α3)×α4, α=α p ×α5=(1.0×α1×α2×α3×α4)×α5, α=α p×α5 = (1.0 ×α1 ×α2 ×α3 ×α4 ×α5) ×α6. Then, the microcomputer 540 finally determines the weighting coefficient α as α = α1 ×α2 ×α3 ×α4 ×α5 ×α6.

[0071] On the other hand, if the microcomputer 540 determines in step S25 that all tables have been selected, i.e., if the weighting coefficient α has been calculated as α = α1 × α2 × α3 × α4 × α5 × α6, the process proceeds from step S25 to step S26. In step S26, the microcomputer 540 performs a process of limiting the final weighting coefficient α to an upper limit value of 1.0 or less, that is, a process of resetting the weighting coefficient α to 1.0 when the weighting coefficient α is greater than 1.0. Figure 20 shows the results of calculating the weighting coefficient α for each distance from the vehicle 100 as α = α1 × α2 × α3 × α4 × α5 × α6, and further shows that the weighting coefficient α is reset to 1.0 in distance regions where the weighting coefficient α exceeds 1.0.

[0072] As described above, if the microcomputer 540 variably sets the weighting coefficient α in the smoothing process in accordance with the distance from the vehicle 100, the speed of the vehicle 100, the curvature of the road, the vibration of the vehicle 100, the period of the smoothing process, and the accuracy of the recognition of the first left-right boundary lines LL1 and LR1, then even if these conditions vary, it is possible to achieve both high levels of convergence to the true value of the second left-right boundary lines LL2 and LR2 after the smoothing process and stability. Note that it is clear that the microcomputer 540 can variably set the weighting coefficient α based on some of the distance from the vehicle 100, the speed of the vehicle 100, the curvature of the road, the vibration of the vehicle 100, the period of the smoothing process, and the accuracy of the recognition of the first left-right boundary lines LL1 and LR1.

[0073] 21 is a block diagram showing a second embodiment of the smoothing process for left and right boundary lines. The target generator 530 of the second embodiment includes a deviation determination processor 530c in addition to the fluctuation suppression processor 530a and storage unit 530b described above.

[0074] The deviation determination processing unit 530c determines the deviation between the information on the first left-right boundary lines LL1, LR1 acquired by the boundary line information acquisition unit 510a and the information on the smoothed second left-right boundary lines LL2, LR2. The deviation between the first left-right boundary lines LL1, LR1 and the second left-right boundary lines LL2, LR2 is the amount of deviation [m] between the left and right boundary lines in the lateral direction of the vehicle 100 (in other words, the lane width direction).

[0075] If there is a deviation between the first left and right boundary lines LL1, LR1 and the second left and right boundary lines LL2, LR2 that exceeds a predetermined threshold, the deviation determination processing unit 530c discards the information about the second left and right boundary lines LL2, LR2 after the smoothing processing and outputs the information about the first left and right boundary lines LL1, LR1 before the smoothing processing as is to the travel route generation unit 610. In other words, if there is a deviation between the first left and right boundary lines LL1, LR1 and the second left and right boundary lines LL2, LR2 that exceeds a predetermined threshold, the deviation determination processing unit 530c cancels the generation of a travel route based on the second left and right boundary lines LL2, LR2 after the smoothing processing and generates a travel route using the information about the first left and right boundary lines LL1, LR1 before the smoothing processing as is.

[0076] 22 is a flowchart showing the flow of a left / right boundary line smoothing program including a deviation determination process executed by the microcomputer 540 of the autonomous driving control unit 500. In step S31 (boundary line information acquisition unit 510a), the microcomputer 540 acquires information about the first left / right boundary lines LL1, LR1 based on the external environment recognition result by the external environment recognition unit 300.

[0077] Next, in step S32 (fluctuation suppression processing unit 530a), the microcomputer 540 performs a smoothing process by weighting and combining the acquired information on the first left-right boundary lines LL1, LR1 and the past information on the second left-right boundary lines LL2p, LR2p, and acquires information on the second left-right boundary lines LL2, LR2 in which fluctuations due to noise, etc. have been suppressed. Note that the microcomputer 540 can variably set the weighting coefficient α used in the smoothing process in step S32 depending on the distance from the vehicle 100, etc., as in the first embodiment.

[0078] Furthermore, in step S33 (deviation determination processing unit 530c), the microcomputer 540 calculates the deviation (shift) between the second left-right boundary lines LL2, LR2 after the smoothing process and the first left-right boundary lines LL1, LR1 before the smoothing process. Then, in step S33, the microcomputer 540 compares the calculated deviation with a predetermined threshold. If the deviation is equal to or less than the predetermined threshold, or if the deviation exceeds the predetermined threshold but has not continued for a predetermined time, the microcomputer 540 proceeds to step S34 and outputs information about the second left-right boundary lines LL2, LR2 after the smoothing process for generating a travel route. On the other hand, if the deviation exceeds the predetermined threshold for a predetermined time, the microcomputer 540 proceeds to step S35 and discards information about the second left-right boundary lines LL2, LR2 after the smoothing process and outputs information about the first left-right boundary lines LL1, LR1 before the smoothing process for generating a travel route.

[0079] The deviation threshold is adapted to determine whether a sudden change in the first left-right boundary lines LL1, LR1 is an amount that requires action for vehicle safety. The predetermined time for determining the duration of the deviation exceeding the threshold is adapted to determine whether the deviation is a temporary deviation caused by noise or the like, or a deviation caused by an obstacle jumping into or protruding from the lane. Obstacles include automobiles, bicycles, pedestrians, and flying objects such as balls.

[0080] In other words, if the deviation is equal to or less than the threshold, the microcomputer 540 determines that the deviation is within an allowable range that ensures vehicle safety, and normally generates a driving route based on the information on the second left-right boundary lines LL2, LR2. On the other hand, if the deviation exceeds the threshold for a predetermined period of time, the microcomputer 540 determines that the deviation is due to an obstacle suddenly appearing, etc., and that vehicle safety may be compromised if a driving route is generated based on the information on the second left-right boundary lines LL2, LR2 that does not sufficiently reflect sudden changes in the first left-right boundary lines LL1, LR1. Then, the microcomputer 540 discards the information on the second left-right boundary lines LL2, LR2, and generates a driving route based on the information on the first left-right boundary lines LL1, LR1.

[0081] The following describes a situation in which the deviation determination processing unit 530c functions. Fig. 23 shows a situation in which an obstacle, such as a ball, suddenly appears in the lane in which the vehicle 100 is traveling. When smoothing processing of the left and right boundary lines is performed, a delay occurs before a sudden change in the actual left and right boundary lines, caused by the obstacle appearing, is reflected in the second left and right boundary lines LL2, LR2 output by the fluctuation suppression processing unit 530a.

[0082] In particular, when the weighting coefficient α is made variable based on the distance from the vehicle 100, a significant delay appears near the vehicle 100 where the degree of smoothing is increased for stabilization. As a result, when an obstacle suddenly appears in front of the vehicle 100, the output of the left and right boundary lines that define the driving range that avoids the obstacle is delayed, which may compromise the safety of the vehicle 100.

[0083] Therefore, when a sudden change in the first left-right boundary lines LL1, LR1 occurs due to, for example, an obstacle appearing in front of the vehicle, and the deviation between the first left-right boundary lines LL1, LR1 and the second left-right boundary lines LL2, LR2 exceeds a threshold value and continues for a predetermined period of time, the deviation determination processing unit 530c discards the information on the second left-right boundary lines LL2, LR2 and instead outputs the information on the first left-right boundary lines LL1, LR1 to the driving path generation unit 610. Here, the first left-right boundary lines LL1, LR1 are boundary line information before smoothing processing, and reflect the appearance of an obstacle in a responsive manner. Therefore, by the deviation determination processing unit 530c outputting the information on the first left-right boundary lines LL1, LR1 instead of the second left-right boundary lines LL2, LR2 to the driving path generation unit 610, delays in setting a drivable range that avoids obstacles can be prevented, and the safety of the vehicle 100 can be ensured.

[0084] The deviation determination process performed by the deviation determination processing unit 530c will be described in detail below with reference to Figures 23 and 24. Figure 24 shows the change in deviation amount with respect to the distance from the vehicle 100 when the obstacle shown in Figure 23 suddenly appears.

[0085] The deviation determination processing unit 530c sets an area of ​​a predetermined width (predetermined width = threshold value × 2) centered on the second left-right boundary line LL2, LR2 as a margin of safety for ensuring vehicle safety. The deviation determination processing unit 530c then determines whether the information on the first left-right boundary line LL1, LR1 before smoothing processing is outside the safety margin area, based on whether the deviation between the first left-right boundary line LL1, LR1 and the second left-right boundary line LL2, LR2 exceeds a predetermined threshold value.

[0086] If the amount of deviation between the first left and right boundary lines LL1, LR1 and the second left and right boundary lines LL2, LR2 exceeds a predetermined threshold for a predetermined period of time, the deviation determination processing unit 530c determines that a sudden change in the first left and right boundary lines LL1, LR1 has occurred due to an obstacle suddenly appearing, etc. Then, the deviation determination processing unit 530c discards the information on the second left and right boundary lines LL2, LR2 and instead outputs information on the first left and right boundary lines LL1, LR1 that reflects the obstacle suddenly appearing, etc., to the travel path generation unit 610.

[0087] On the other hand, if the amount of deviation between the first left and right boundary lines LL1, LR1 and the second left and right boundary lines LL2, LR2 does not exceed the predetermined threshold, the deviation determination processing unit 530c outputs information about the second left and right boundary lines LL2, LR2 to the travel route generation unit 610. In addition, if the amount of deviation exceeds the predetermined threshold but the duration does not reach the predetermined time, the deviation determination processing unit 530c also outputs information about the second left and right boundary lines LL2, LR2 to the travel route generation unit 610.

[0088] In this way, the deviation determination processing unit 530c calculates the deviation amount, which is the amount of deviation in the left-right direction of the first left-right boundary lines LL1, LR1 from the second left-right boundary lines LL2, LR2, for each distance from the vehicle 100. If the calculated deviation amount remains greater than a threshold value set as a safety margin for a predetermined period of time, the deviation determination processing unit 530c determines that the deviation amount has increased due to a sudden change in the first left-right boundary lines LL1, LR1 due to, for example, an obstacle suddenly appearing (see FIG. 24). If the large deviation amount continues, the deviation determination processing unit 530c discards the information on the second left-right boundary lines LL2, LR2 and outputs the information on the first left-right boundary lines LL1, LR1 to the driving path generation unit 610, thereby causing the driving path generation unit 610 to generate a driving path based on a drivable range that reflects sudden changes such as an obstacle suddenly appearing, thereby reducing control delays due to, for example, an obstacle suddenly appearing.

[0089] Here, the deviation determination processing unit 530c can vary the deviation threshold (in other words, the width of the safety margin) depending on the distance from the vehicle 100. Fig. 25 shows the correlation between the distance from the vehicle 100 and the deviation threshold.

[0090] The information about the first left-right boundary lines LL1, LR1 acquired by the boundary line information acquisition unit 510a tends to fluctuate more as the distance from the vehicle 100 increases. Therefore, the deviation determination processing unit 530c sets a larger threshold value as the distance from the vehicle 100 increases, thereby making it possible to distinguish between fluctuations in the first left-right boundary lines LL1, LR1 due to noise or the like and sudden changes in the first left-right boundary lines LL1, LR1 due to an obstacle appearing in front of the vehicle 100.

[0091] The deviation determination processing unit 530c can change the length of the information about the first left and right boundary lines LL1, LR1 to be output to the driving path generation unit 610, depending on the boundary line length at which a deviation exceeding the threshold has occurred. That is, the deviation determination processing unit 530c can replace only the portion at which a deviation exceeding the threshold has occurred with the information about the first left and right boundary lines LL1, LR1, and further, if the length of the portion at which a deviation exceeding the threshold has occurred exceeds a set value, can discard the second left and right boundary lines LL2, LR2 and replace the entire portion with the first left and right boundary lines LL1, LR1.

[0092] It is assumed that sudden changes in the first left and right boundary lines LL1, LR1 are caused by pedestrians, bicycles, automobiles, etc. jumping in or out of the lane. For this reason, the threshold for determining whether the deviation is long or short can be set based on the total length (for example, about 12 m) of a large truck, which is one of the longest obstacles that jump out.

[0093] Here, the deviation determination processing unit 530c can use the ratio of the boundary line length where a deviation exceeding a threshold has occurred to the entire length of the first left and right boundary lines LL1, LR1 (the distance from the vehicle 100 to the farthest position where the first left and right boundary lines LL1, LR1 are recognized) as an index when determining whether to replace only the deviation portion with the first left and right boundary lines LL1, LR1 or to replace the entire boundary line with the first left and right boundary lines LL1, LR1. Figure 26 illustrates an example of the entire length of the left and right boundary lines and the length of the portion where a deviation exceeding the threshold has occurred (deviation range).

[0094] The deviation determination processing unit 530c calculates the ratio of the length of the deviation range to the entire length of the left and right boundary lines, and if this ratio is equal to or less than a threshold, i.e., if the ratio of the deviation range to the entire length of the left and right boundary lines is equal to or less than a predetermined ratio, it replaces only the deviation portion with the first left and right boundary lines LL1, LR1 and outputs the rest of the deviation portion as the second left and right boundary lines LL2, LR2. On the other hand, if this ratio exceeds the threshold, i.e., if the ratio of the deviation range to the entire length of the left and right boundary lines exceeds a predetermined ratio, the deviation determination processing unit 530c discards the second left and right boundary lines LL2, LR2 and replaces the entire length, including the deviation portion, with the first left and right boundary lines LL1, LR1. In this way, by switching between replacing only the deviation portion with the first left and right boundary lines LL1, LR1 or replacing the entire portion with the first left and right boundary lines LL1, LR1 depending on the ratio of the length of the deviation range to the total length of the left and right boundary lines, when the ratio is relatively small, the second left and right boundary lines LL2, LR2 after the smoothing process can be effectively utilized.

[0095] 27 illustrates a situation in which partial replacement (replacement of only the deviation portion) and full replacement are switched depending on the ratio of the length of the deviation range to the entire length of the left and right boundary lines. In FIG. 27, an obstacle ball is projected into the lane of vehicle 100. In the upper diagram, the deviation range is relatively short, whereas in the lower diagram, the deviation range is longer than in the upper diagram.

[0096] In this case, if the ratio of the length of the deviation range to the entire length of the left and right boundary lines is equal to or less than the threshold value in the case of the upper part of Fig. 27, only the deviation portion is replaced with the first left and right boundary lines LL1, LR1. On the other hand, in the case of the lower part of Fig. 27, if the ratio of the length of the deviation range to the entire length of the left and right boundary lines exceeds the threshold value, the second left and right boundary lines LL2, LR2 are discarded, and the entire boundary line including the deviation portion is replaced with the first left and right boundary lines LL1, LR1.

[0097] The technical ideas described in the above embodiments can be used in appropriate combinations as long as no contradictions arise. Furthermore, although the contents of the present invention have been specifically described with reference to preferred embodiments, it is obvious that a person skilled in the art can adopt various modified embodiments based on the basic technical ideas and teachings of the present invention.

[0098] For example, the deviation determination processing unit 530c can variably set the deviation determination threshold based on the lane width defined by the second left and right boundary lines LL2, LR2, the distance between the travel path of the vehicle 100 and the second left and right boundary lines LL2, LR2, etc. In particular, the deviation determination processing unit 530c can reduce the deviation determination threshold as the lane width (width of the driving range) defined by the second left and right boundary lines LL2, LR2 becomes narrower, and can also reduce the deviation determination threshold as the distance between the travel path of the vehicle 100 and the second left and right boundary lines LL2, LR2 becomes narrower. In other words, the deviation determination processing unit 530c can reduce the threshold as the distance from the second left and right boundary lines LL2, LR2 to the vehicle 100 becomes narrower and an object running into the lane, for example, is more likely to interfere with the vehicle 100, thereby making it easier to replace the boundary line with the first left and right boundary lines LL1, LR1.

[0099] Furthermore, the present invention is not limited to a configuration in which the weighting coefficient α1 is corrected by the correction values ​​α2, α3, α4, α5, and α6 to determine the final weighting coefficient α, but may be a process in which the weighting coefficient α is changed in accordance with the distance from the vehicle 100, the speed of the vehicle 100, the curvature of the road, the vibration of the vehicle 100, the period of the smoothing process, and the accuracy of recognition of the first left and right boundary lines LL1 and LR1. Furthermore, the process in which the weighting coefficient α1 and the correction values ​​α2, α3, α4, α5, and α6 are determined is not limited to a conversion process using a table, but may be a calculation process using a preset function.

[0100] Furthermore, the driving route generation unit 610 can generate the center of the driving range indicated by the left and right boundary lines as the driving route for the vehicle 100, and is not limited to planning a driving route that can suppress the acceleration and jerk of the vehicle 100. Furthermore, the vehicle control system 200 can be configured to include one control unit that combines the control functions of the autonomous driving control unit 500 and the vehicle motion control unit 600, and further, the vehicle control device of the present invention can also be configured with three or more control units (microcomputers).

[0101] 100...vehicle, 200...vehicle control system, 300...external environment recognition unit, 500...autonomous driving control unit, 510...surrounding situation recognition unit, 510a...boundary line information acquisition unit, 520...action planning unit, 530...target generation unit, 530a...variation suppression processing unit, 530b...storage unit, 530c...deviation determination processing unit, 540...microcomputer (control unit), 600...vehicle motion control unit, 610...travel path generation unit, 620...following control unit, 630...microcomputer (control unit), 700...actuator unit

Claims

1. A vehicle control device mounted on a vehicle, comprising: a boundary line information acquisition unit that acquires information on a first left and right boundary line indicating a drivable range in a travel route on which the vehicle travels; a variation suppression processing unit that executes smoothing processing on the first left and right boundary line to obtain information on a second left and right boundary line after the smoothing processing; a storage unit that stores the information on the second left and right boundary line; and a travel route generation unit that obtains a travel route on which the vehicle travels using the information on the second left and right boundary line. The variation suppression processing unit executes the smoothing processing by weighting and synthesizing the information on the first left and right boundary line and the information on the past second left and right boundary line stored in the storage unit, respectively. Vehicle control device.

2. The vehicle control device according to claim 1, further comprising a deviation determination processing unit that determines a deviation between the information on the first left and right boundary line and the information on the second left and right boundary line, and outputs the information on the first left and right boundary line to the travel route generation unit when it is determined that there is a deviation exceeding a predetermined threshold. Vehicle control device.

3. The vehicle control device according to claim 1, wherein the variation suppression processing unit executes smoothing processing by weighting and synthesizing the previous value included in the information on the past second left and right boundary line and the nearest neighbor point closest to the previous value among the information on the first left and right boundary line, respectively. Vehicle control device.

4. The vehicle control device according to claim 1, wherein the variation suppression processing unit changes the weighting based on at least one of the distance from the vehicle, the speed of the vehicle, the curvature of the travel route, the vibration of the vehicle, the period of the smoothing processing, or the accuracy of recognition of the first left and right boundary line. Vehicle control device.

5. The vehicle control device according to claim 2, wherein the deviation determination processing unit outputs the information on the first left and right boundary line to the travel route generation unit when a state in which the deviation between the information on the first left and right boundary line and the information on the second left and right boundary line exceeds the threshold continues for a predetermined time. Vehicle control device.

6. The vehicle control device according to claim 5, wherein the deviation determination processing unit changes the threshold according to the distance from the vehicle. Vehicle control device.

7. The vehicle control device according to claim 2, wherein the deviation determination processing unit changes the length of the information of the first left and right boundary lines output to the travel route generation unit according to the length of the boundary line at which a deviation exceeding the threshold value has occurred. Vehicle control device.

8. A vehicle control method executed by a control unit mounted on a vehicle, the method comprising: obtaining information on first left and right boundary lines indicating a travelable range on a travel route along which the vehicle travels; performing a smoothing process on the first left and right boundary lines to obtain information on second left and right boundary lines after the smoothing process; storing the information on the second left and right boundary lines; and obtaining a travel route along which the vehicle travels using the information on the second left and right boundary lines. The step of obtaining the information on the second left and right boundary lines includes performing the smoothing process by weighting and synthesizing the information on the first left and right boundary lines and the stored information on the past second left and right boundary lines. Vehicle control method.