Travel control method and travel control device

By detecting lane shape and applying curvature constraints, the method generates a smooth target travel trajectory that aligns with the occupant's intentions, addressing the issue of non-smooth trajectories in existing vehicle path prediction systems.

JP2024170092APending Publication Date: 2024-12-06NISSAN MOTOR CO LTD
View PDF 1 Cites 0 Cited by

Patent Information

Application Number
JP2023087059
Authority / Receiving Office
JP · JP
Patent Type
Applications
Current Assignee / Owner
Filing Date
2023-05-26
Publication Date
2024-12-06

AI Technical Summary

Technical Problem

Existing methods for generating target travel trajectories based on past vehicle travel data often result in jagged or non-smooth trajectories, failing to accurately reflect the intended path of the vehicle occupant.

Method used

The method involves detecting the lane shape in front of the vehicle, generating a first target travel trajectory based on past travel data, setting a driving area, and creating a second target trajectory that satisfies curvature and curvature change constraints to ensure a smooth path.

Benefits of technology

This approach generates a smoother target travel trajectory that aligns with the occupant's intentions by utilizing past travel data while adhering to preset constraints, enhancing the vehicle's control and alignment with the intended path.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure 2024170092000001_ABST
    Figure 2024170092000001_ABST
Patent Text Reader

Abstract

To generate a smoother target travel track, while taking an advantage of the target travel track generated based on a travel track on which a vehicle has traveled in the past.SOLUTION: A travel control method includes: detecting a shape of a lane in front of an own vehicle of a travel lane on which the own vehicle is traveling (S1); generating a first target travel track on the travel lane in front of the own vehicle, on the basis of the travel track on which a vehicle has traveled in the past (S2); setting a travel area in which the own vehicle travels on the travel lane in front of the own vehicle, on the basis of the first target travel track (S3); generating a second target travel track on which the vehicle travels in the travel area so as to satisfy a constraint condition predetermined for at least one of a curvature of the second target travel track and a curvature change (S4); and controlling the own vehicle so that the vehicle travels along the second target track (S5).SELECTED DRAWING: Figure 4
Need to check novelty before this filing date? Find Prior Art

Description

[Technical field]

[0001] The present invention relates to a cruise control method and a cruise control device. [Background technology]

[0002] Patent Document 1 listed below proposes a vehicle driving control device that learns driving trajectories during past manual driving and generates a target trajectory during automatic driving. [Prior art documents] [Patent documents]

[0003] [Patent Document 1] JP 2018-022353 A Summary of the Invention [Problem to be solved by the invention]

[0004] By generating a target driving trajectory of the vehicle using a learning model that has learned the driving trajectory of the vehicle driven manually in the past, it is possible to generate a target driving trajectory that is close to the trajectory intended by the occupant. However, the target driving trajectory generated using the learning model is not necessarily a smooth trajectory. The present invention has an object to generate a smoother target driving trajectory while taking advantage of the advantages of a target driving trajectory generated based on a driving locus traveled by a vehicle in the past. [Means for solving the problem]

[0005] In a driving control method according to one embodiment of the present invention, the shape of the lane ahead of the vehicle in the driving lane in which the vehicle is driving is detected, a first target driving trajectory in the driving lane ahead of the vehicle is generated based on a driving trajectory traveled by the vehicle in the past, a driving area in which the vehicle will drive in the driving lane ahead of the vehicle is set based on the first target driving trajectory, a second target driving trajectory for driving within the driving area is generated so as to satisfy predetermined constraints for at least one of the curvature and curvature change of the second target driving trajectory, and the vehicle is controlled to drive along the second target driving trajectory. Effect of the Invention

[0006] According to the present invention, it is possible to generate a smoother target driving trajectory while taking advantage of the advantages of a target driving trajectory generated based on a driving locus traveled by a vehicle in the past. [Brief description of the drawings]

[0007] [Figure 1] 1 is a schematic configuration diagram of an example of a driving control device according to an embodiment; [Diagram 2] FIG. 2 is a block diagram illustrating an example of a functional configuration of a controller according to the first embodiment. [Diagram 3] FIG. 2A is a schematic diagram of an example of a first target driving trajectory, and FIG. 2B is a schematic diagram of an example of a driving region and a second target driving trajectory. [Figure 4] 4 is a flowchart of an example of a driving control method according to the first embodiment. [Diagram 5] FIG. 11 is a block diagram illustrating an example of a functional configuration of a controller according to a second embodiment. [Figure 6] FIG. 4A is a schematic diagram illustrating an example of learning data of a learning model, and FIG. 4B is a schematic diagram illustrating an example of data on the shape of a driving lane and a first target driving trajectory. [Figure 7] 13(a) to 13(c) are schematic diagrams illustrating an example of a driving control method according to a second embodiment. [Figure 8] 10 is a flowchart of an example of a driving control method according to a second embodiment. [Figure 9]FIG. 13 is a schematic diagram illustrating an example of a driving control method according to a third embodiment. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

[0008] Hereinafter, an embodiment of the present invention will be described with reference to the drawings. Note that each drawing is a schematic view and may differ from the actual one. In addition, the embodiment of the present invention shown below is an example of an apparatus and a method for embodying the technical idea of ​​the present invention, and the technical idea of ​​the present invention does not specify the structure, arrangement, etc. of the components as described below. The technical idea of ​​the present invention can be modified in various ways within the technical scope defined by the claims described in the claims.

[0009] (First embodiment) (composition) FIG. 1 is a schematic configuration diagram of an example of a cruise control device according to an embodiment. The host vehicle 1 includes a cruise control device 10 that controls the traveling of the host vehicle 1. The cruise control by the cruise control device 10 includes automatic control that sets a target traveling trajectory based on the shape of the lane on which the host vehicle 1 travels, and controls a steering mechanism so that the host vehicle 1 travels along the target traveling trajectory. For example, the cruise control by the cruise control device 10 may be autonomous driving control that automatically drives the host vehicle 1 without the involvement of a driver, based on the traveling environment around the host vehicle 1.

[0010] The driving control device 10 includes an external sensor 11, a vehicle sensor 12, a positioning device 13, a map database (map DB) 14, a navigation device 15, an actuator 17, and a controller 18. The external sensor 11 includes a plurality of different types of object detection sensors mounted on the vehicle 1, such as a laser radar, a millimeter wave radar, a camera, and a LIDAR (Light Detection and Ranging, Laser Imaging Detection and Ranging) that detect objects around the vehicle 1.

[0011] The vehicle sensor 12 is mounted on the host vehicle 1 and detects various information (vehicle signals) obtained from the host vehicle 1. The vehicle sensor 12 includes, for example, a vehicle speed sensor that detects the vehicle speed of the host vehicle 1, a wheel speed sensor that detects the rotational speed of the tires of the host vehicle 1, an acceleration sensor that detects the lateral acceleration Gy and longitudinal acceleration Gx of the host vehicle 1, a steering angle sensor that detects the steering angle of the steering wheel, a turning angle sensor that detects the turning angle of the steered wheels, a gyro sensor that detects the angular velocity of the host vehicle 1, a yaw rate sensor that detects the yaw rate, an accelerator sensor that detects the accelerator opening of the host vehicle, and a brake sensor that detects the amount of brake operation by the driver.

[0012] The positioning device 13 includes a Global Navigation System (GNSS) receiver and receives radio waves from multiple navigation satellites to measure the current position of the vehicle 1. The GNSS receiver may be, for example, a Global Positioning System (GPS) receiver. The positioning device 13 may be, for example, an inertial navigation system. The map database 14 stores road map data. For example, the map database 14 may store high-precision map data (hereinafter simply referred to as "high-precision map") suitable as map information for autonomous driving. The map database 14 may also store map data for navigation (hereinafter simply referred to as "navigation map").

[0013] The navigation device 15 recognizes the current position of the vehicle 1 using the positioning device 13, and obtains map information for the current position from the map database 14. The navigation device 15 sets a target route to the destination input by the occupant, and provides route guidance to the occupant along this target route. The navigation device 15 also outputs information on the set target route to the controller 18. When executing autonomous driving control, the controller 18 automatically drives the vehicle 1 so that it travels along the target route set by the navigation device 15.

[0014] The actuator 17 operates the steering wheel, accelerator opening and braking device of the host vehicle in response to control signals from the controller 18 to generate vehicle behavior of the host vehicle. The actuator 17 includes a steering actuator, an accelerator opening actuator and a brake control actuator. The steering actuator controls the steering direction and steering amount of the host vehicle. The accelerator opening actuator controls the accelerator opening of the host vehicle. The brake control actuator controls the braking operation of the braking device of the host vehicle.

[0015] The controller 18 is an electronic control unit that controls the running of the host vehicle 1. The controller 18 includes a processor 18a and peripheral components such as a storage device 18b. The processor 18a may be, for example, a CPU or an MPU. The storage device 18b may include a semiconductor storage device, a magnetic storage device, an optical storage device, etc. The storage device 18b may include memories such as a register, a cache memory, and a ROM and a RAM used as a main storage device.

[0016] The functions of the controller 18 described below are realized, for example, by the processor 18a executing a computer program stored in the storage device 18b. The controller 18 may be formed of dedicated hardware for executing each information processing described below. For example, the controller 18 may include a functional logic circuit set in a general-purpose semiconductor integrated circuit. For example, the controller 18 may include a PLD such as an FPGA.

[0017] 2 is a block diagram of an example of a functional configuration of the controller 18 of the first embodiment. The controller 18 includes a lane boundary detection unit 30, a first target driving trajectory generation unit 31, a driving area setting unit 32, a second target driving trajectory generation unit 33, and a vehicle control unit 34. The lane boundary detection unit 30 detects the shape of the lane ahead of the host vehicle 1 in the driving lane in which the host vehicle 1 is traveling.

[0018] See Fig. 3(a). For example, the lane boundary detection unit 30 may detect the positions of lane boundary points PR0-PR8 on the right lane boundary line and the positions of lane boundary points PL0-PL8 on the left lane boundary line as the lane shape of the driving lane ahead of the vehicle 1. The lane boundary detection unit 30 may detect the positions of the lane boundary points PR0-PR8 and PL0-PL8 based on, for example, an image captured by the camera of the external sensor 11 or distance measurement data by a distance measurement device such as a laser radar, a millimeter wave radar, or a LIDAR. Note that the number of lane boundary points on the lane boundary lines detected by the lane boundary detection unit 30 is not limited to eight on each side, and may be more or less than eight.

[0019] Referring to Fig. 2, the first target driving trajectory generating unit 31 generates a first target driving trajectory T1 for the driving lane ahead of the host vehicle 1. For example, the first target driving trajectory generating unit 31 may have a learning model 31a that learns the lane shape, vehicle speed, lateral acceleration and longitudinal acceleration of the vehicle, and driving trajectory of a lane on which the vehicle has been manually driven in the past.

[0020] For example, the learning model 31a may be a multi-layered neural network including at least an input layer and an output layer. For example, when the lane shape of the driving lane detected by the lane boundary detection unit 30 and the vehicle speed, lateral acceleration Gy, and vertical acceleration Gx of the host vehicle 1 detected by the vehicle sensor 12 are input to the input layer, the learning model 31a outputs the learned driving trajectory as the first target driving trajectory T1 from the output layer. For example, the input layer receives the X coordinates (coordinates in the lane direction) and Y coordinates (coordinates in the lane width direction) of 30 left and right lane boundary points detected at approximately equal intervals in a predetermined section (e.g., 70 m) ahead of the vehicle 1, as well as the vehicle speed, lateral acceleration Gy, and longitudinal acceleration Gx. The output layer outputs the X coordinates and Y coordinates of 30 passing points at approximately equal intervals on the travel trajectory in the 70 m section of the travel lane ahead of the vehicle 1.

[0021] When generating a driving trajectory using such a learning model 31a, a problem may occur in which the driving trajectory becomes jagged, particularly when driving a route that has not been driven before, or when the route has been driven before but the input at the time of learning is significantly different from the input at the time of driving. In addition, the past manual driving data (lane shape, vehicle speed, lateral acceleration, vertical acceleration, driving trajectory) to be learned by the learning model 31a may basically be driving data of the vehicle 1 for each specified section (e.g., 70 m) of a curved road having a curvature equal to or greater than a specified value.

[0022] For example, the first target driving trajectory generating unit 31 may generate a driving trajectory on which the vehicle has traveled in the past as the first target driving trajectory T1. The driving trajectory on which the vehicle has traveled in the past may be, for example, a driving trajectory on which the vehicle has traveled in a case where the vehicle has traveled by manual driving or a driving trajectory on which the vehicle has traveled by driving assistance that assists in steering control of the vehicle.

[0023] The dotted line Ta in Figure 3(a) indicates the vehicle's driving trajectory learned by the learning model 31a when the vehicle was driven manually in the past, and the solid line T1 connecting the circular plots P0, P1a to P8a represents the first target driving trajectory formed in front of the vehicle 1 with the current position P0 of the vehicle 1 as the base point. The data on lane shape and driving trajectory used as learning data for the learning model 31a may be data on the lane shape and driving trajectory when the host vehicle 1 is driven manually, or may be data on the lane shape and driving trajectory when a vehicle other than the host vehicle 1 is driven manually.

[0024] 2, the driving area setting unit 32 sets a driving area Rt in which the host vehicle 1 will travel on the driving lane ahead of the host vehicle 1, based on the first target driving trajectory T1 set by the first target driving trajectory generating unit 31. Refer to Figure 3(b). The hatched area is a schematic representation of the travel area Rt. The dashed lines BR and BL indicate the right boundary line BR, which is the boundary line on the right side of the travel area Rt, and the left boundary line RL, which is the boundary line on the left side of the travel area Rt, respectively.

[0025] For example, the driving area setting unit 32 may set a sequence of points forming the right boundary line BR and the left boundary line RL by alternately allocating the points P0, P1a to P8a on the first target driving trajectory T1 to the right boundary line BR and the left boundary line RL. Also, for example, the driving area setting unit 32 may set a sequence of points forming the right boundary line BR and the left boundary line RL by allocating the points P0, P1a to P8a on the first target driving trajectory T1 located to the right of the actual driving trajectory Ta to the right boundary line BR and the points located to the right to the left boundary line RL.

[0026] The driving area setting unit 32 may generate the right boundary line BR based on the points on the first target driving trajectory T1 assigned to the right boundary line BR, and may generate the left boundary line RL based on the points on the first target driving trajectory T1 assigned to the left boundary line RL. For example, the right boundary line BR may be generated by connecting the points assigned to the right boundary line BR with straight lines, and the left boundary line RL may be generated by connecting the points assigned to the left boundary line RL with straight lines. Alternatively, the right boundary line BR may be generated by forming an approximation curve or an approximation straight line based on the points assigned to the right boundary line BR, and the left boundary line RL may be generated by forming an approximation curve or an approximation straight line based on the points assigned to the left boundary line RL.

[0027] 2, the second target running trajectory generating unit 33 generates the second target running trajectory T2 within the running area Rt set by the running area setting unit 32. At this time, the second target running trajectory generating unit 33 generates the second target running trajectory T2 such that at least one of the curvature and the curvature change of the second target running trajectory T2 satisfies a preset constraint condition. The dashed dotted line in FIG. 3(b) represents a second target driving trajectory T2 generated within the driving region Rt. The vehicle control unit 34 generates a control signal for the actuator 17 so that the host vehicle 1 travels along the second target travel trajectory T2, and outputs the generated control signal to the actuator 17.

[0028] In this way, by generating the second target driving trajectory T2 within the driving region Rt set based on the first target driving trajectory T1, the advantages of the first target driving trajectory T1 generated by the learning model 31a that learned the driving trajectory Ta that the vehicle traveled in the past under manual driving can be utilized for the second target driving trajectory T2. For example, a target driving trajectory close to the trajectory intended by the occupant can be generated. Furthermore, by generating the second target running trajectory T2 such that at least one of the curvature and the curvature change of the second target running trajectory T2 satisfies a preset constraint condition, a smooth second target running trajectory T2 can be generated.

[0029] (operation) FIG. 4 is a flowchart of an example of the driving control method according to the first embodiment. In step S1, the lane boundary detection unit 30 detects the shape of the lane ahead of the host vehicle 1 in the driving lane in which the host vehicle 1 is traveling. In step S2, the first target driving trajectory generation unit 31 generates a first target driving trajectory T1 by inputting the lane shape of the driving lane detected by the lane boundary detection unit 30 into a learning model 31a that has learned the lane shape and driving trajectory of lanes on which the vehicle has previously driven manually.

[0030] In step S3, the traveling area setting unit 32 sets the traveling area Rt based on the first target traveling trajectory T1. In step S4, the second target running trajectory generating unit 33 generates a second target running trajectory T2 for running within the running area Rt so as to satisfy a preset constraint condition for at least one of the curvature and the curvature change of the second target running trajectory T2. In step S5, the vehicle control unit 34 controls the host vehicle 1 to travel along the second target travel trajectory T2, after which the process ends.

[0031] Second embodiment 5 is a block diagram of an example of the functional configuration of the controller 18 of the second embodiment. The controller 18 of the second embodiment has a similar configuration to the controller 18 of the first embodiment, and the same or similar components are given the same reference symbols and will not be described again. The first target traveling trajectory generating unit 31 of the second embodiment includes a learning model 31a and a predicted trajectory calculating unit 31b.

[0032] As described above, the learning model 31a is a learning model that learns the lane shapes and driving trajectories of lanes on which the vehicle has been manually driven in the past. The learning model 31a may be configured, for example, by a neural network, a random forest, a support vector machine, or the like. As learning data for making the learning model 31a learn the lane shape, position data of lane boundary points on lane boundary lines may be used, for example, similar to the lane boundary points PR0 to PR8 and PL0 to PL8 shown in FIG. 3(a).

[0033] 6(a) is a schematic diagram illustrating an example of learning data of the learning model 31a. For example, when the vehicle V was manually driven in the past, a section of a predetermined length of the lane on which the vehicle V traveled is divided into N divided sections (N is an integer of 2 or more). 6, the division number N is "8", and the lane ahead of the vehicle V is divided into eight divided sections S1 to S8. The division number N is not limited to "8", and may be more or less than "8".

[0034] Then, the position data of lane boundary points p00 and p10 on the right and left lane boundary lines at the start point of divided section S1, and the position data of lane boundary points p01 to p08 and p11 to p18 on the right and left lane boundary lines at the end points of each of divided sections S1 to S8 are trained in learning model 31a as learning data for the vehicle alignment shape.

[0035] On the other hand, as learning data for the driving trajectory Ta, the learning model 31a may learn position data of the intersections between the driving trajectory Ta and the section line Ls, which is a straight line connecting the right-side lane boundary points p00 to p08 and the left-side lane boundary points p10 to p18, respectively. The section line Ls is a straight line extending in the lane width direction, and separates adjacent divided sections from among the divided sections S1 to S8.

[0036] For example, as position data of the intersection between the section line Ls and the driving trajectory Ta, the learning model 31a may learn the ratio (Ln / Ld) of the distance Ln from the left lane boundary line to the intersection point to the total length of the section line Ls (i.e., the distance between the right and left lane boundary lines) Ld. If the intersection of the section line Ls and the driving trajectory Ta is located on the left lane boundary line, the ratio (Ln / Ld) is "0", and if it is located on the right lane boundary line, the ratio (Ln / Ld) is "1".

[0037] 5, the lane boundary detection unit 30 detects the shape of the lane ahead of the host vehicle 1 in which the host vehicle 1 is traveling, and outputs data on the vehicle line shape to the first target traveling trajectory generation unit 31 and the traveling area setting unit 32. For example, the lane boundary detection unit 30 may generate data on the vehicle line shape of the driving lane in which the host vehicle 1 is traveling in the same format as the learning data of the vehicle line shape that is to be learned by the learning model 31a. See Fig. 6(b). For example, the lane boundary detection unit 30 may divide the lane ranging from the current position P0 of the vehicle 1 to a point a predetermined length ahead into N divided sections (8 in the example of Fig. 6(b)).

[0038] The lane boundary detection unit 30 may output the positions of lane boundary points PR0 and RL0 on the right and left lane boundary lines at the current position P0 of the vehicle 1, and the position data of lane boundary points PR1 to PR8 and PL1 to PL8 on the right and left lane boundary lines at the end points of each divided section, as lane shape data for the driving lane. The thin solid line Ls in Figure 6(b) is a section line Ls that connects the right-side lane boundary points PR0 to PR8 with the left-side lane boundary points PL0 to PL8 with straight lines, and the section line Ls separates adjacent divided sections among the N divided sections.

[0039] The first target driving trajectory generating unit 31 inputs the data on the vehicle line shape output from the lane boundary detecting unit 30 to a learning model 31a. Based on the input data on the lane shape, the learning model 31a generates a first target driving trajectory T1 for the driving lane ahead of the host vehicle 1. A solid line T1 in Fig. 6(b) indicates the first target driving trajectory T1.

[0040] For example, the learning model 31a may output the position of the first target driving trajectory T1 on each section line Ls as shown by a circle plot in FIG. 6B as data representing the first target driving trajectory T1. That is, the learning model 31a may output the intersection point between each section line Ls and the first target driving trajectory T1. For example, the learning model 31a may output the ratio (Ln / Ld) of the distance Ln from the left lane boundary line to the intersection point to the total length of the section line Ls (i.e., the distance between the right and left lane boundary lines) Ld as data representing the intersection point between the section line Ls and the first target driving trajectory T1. If the intersection of the section line Ls and the first target driving trajectory T1 is located on the left lane boundary line, the ratio (Ln / Ld) is "0", and if it is located on the right lane boundary line, the ratio (Ln / Ld) is "1".

[0041] Please refer to Fig. 5. The predicted trajectory calculation unit 31b calculates the positions of the intersections between each section line Ls and the first target driving trajectory T1 based on the position information of the lane boundary points PR0 to PR8 and the left lane boundary points PL0 to PL8 detected by the lane boundary detection unit 30 and the data of the ratio (Ln / Ld) output from the learning model 31a, and generates position data of the sequence of the intersections as data of the first target driving trajectory T1.

[0042] The driving area setting unit 32 sets a driving area Rt in which the host vehicle 1 travels in the driving lane ahead of the host vehicle 1, based on the first target driving trajectory T1 set by the first target driving trajectory generating unit 31. The driving area setting unit 32 of the second embodiment includes an area width setting unit 32a and an area boundary setting unit 32b. The region width setting unit 32a sets the travel region width W, which is the width of the travel region Rt, based on the shape of the first target travel trajectory T1. For example, the region width setting unit 32a may set the travel region width W based on the smoothness of the first target travel trajectory T1. For example, the travel region width W may be larger when the shape of the first target travel trajectory T1 is not smooth compared to the travel region width W when the shape of the first target travel trajectory T1 is smoother.

[0043] For example, the area width setting unit 32a may select a plurality of points on the first target running trajectory T1 and generate an approximation line of these plurality of points. Then, the running area width W may be set based on the distance from the points on the first target running trajectory T1 to the approximation line. For example, the running area width W may be set based on a statistical value (maximum value, average value, median value, etc.) of the distance from the points on the first target running trajectory T1 to the approximation line.

[0044] FIG. 7(a) is a schematic diagram illustrating an example of a method for setting the travel area width W based on an approximation line of a plurality of points on the first target travel trajectory T1. For example, the region width setting unit 32a sets a plurality of groups each including three adjacent points on the first target running trajectory T1. In the following description, the points on the first target running trajectory T1 may be referred to as "trajectory points." In FIG. 7(a), for the sake of simplicity, only a group consisting of trajectory points P0, P1a, and P2a and a group consisting of P1a, P2a, and P3a are illustrated among the plurality of groups of trajectory points, but other groups are set in the same manner.

[0045] The area width setting unit 32a calculates a straight line connecting the first point and the third point at both ends of three adjacent trajectory points, and calculates the distance from a second point sandwiched between the first point and the third point to the straight line connecting the first point and the third point. In the example of FIG. 7(a), for a group consisting of trajectory points P0, P1a, and P2a, a straight line La1 connecting the first point P0 and the third point P2a is calculated, and a distance D1 from the second point P1a to the straight line La1 is calculated.

[0046] For the group consisting of the trajectory points P1a, P2a, and P3a, a straight line La2 is calculated connecting the first point P1a and the third point P3a, and a distance D2 is calculated from the second point P2a to the straight line La2. Similarly, for the other trajectory point groups, the distances from the second point to the straight line connecting the first point and the third point are calculated. The area width setting unit 32a sets the travel area width W based on distance statistics (maximum, average, median, etc.) calculated for each group of multiple trajectory points.

[0047] The area boundary setting unit 32b sets a driving area Rt having a driving area width W in the lane width direction based on the driving area width W set by the area width setting unit 32a and the first target driving trajectory T1. FIG. 7B is a schematic diagram illustrating an example of a method for setting boundary points that form the right boundary line BR and the left boundary line BL of the traveling area Rt. First, the area boundary setting unit 32b sets each of the trajectory points P1a to P8a (circle plots) on the first target driving trajectory T1 as the first boundary point, which is either one of the boundary points that define the boundaries at both ends in the lane width direction of the driving area Rt (i.e., the boundary point that forms the right boundary line BR and the boundary point that forms the left boundary line BL).

[0048] Next, the area boundary setting unit 32b sets a group of at least three trajectory points (circle plots) on the first target traveling trajectory T1, and generates an approximation straight line (dashed line) of at least three trajectory points included in the group. The area boundary setting unit 32b sets a point (square plot) obtained by moving each of the at least three trajectory points in each direction toward the approximation line from these trajectory points along the lane width direction by the driving area width W as a second boundary point, which is the other of the boundary points that define the boundaries at both ends of the lane width direction of the driving area Rt.

[0049] In the example of Fig. 7(b), for a group of three trajectory points P0, P1a, and P2a, an approximation line Lb1 of the trajectory points P0, P1a, and P2a is calculated. For example, the area boundary setting unit 32b may calculate the approximation line Lb1 based on the least squares method. The area boundary setting unit 32b sets a point P1b (square plot) obtained by moving the trajectory point P1a in the direction from the trajectory point P1a toward the approximation line Lb1 by the driving area width W along the lane width direction (i.e., along the section line Ls) as the second boundary point. The area boundary setting unit 32b also sets a point P2b, which is obtained by moving the trajectory point P2a by the driving area width W along the lane width direction in the direction from the trajectory point P2a toward the approximate line Lb1, as a second boundary point. Regarding the current position P0 of the vehicle 1, points P0b and P0c, which are obtained by moving the current position P0 to the left and right along the lane width direction by the driving area width W / 2, may be set as the first and second boundary points at both ends of the lane width direction of the driving area Rt.

[0050] Similarly, for a group of three trajectory points P3a, P4a, and P5a, an approximation line Lb2 of the trajectory points P3a, P4a, and P5a is calculated. The area boundary setting unit 32b sets a point P3b, which is obtained by moving the trajectory point P3a by the driving area width W along the lane width direction in the direction from the trajectory point P3a toward the approximation line Lb2, as a second boundary point.

[0051] A point P4b obtained by moving the trajectory point P4a from the trajectory point P4a in the direction toward the approximation line Lb2 along the lane width direction by the driving region width W is set as a second boundary point. A point P5b obtained by moving the trajectory point P5a from the trajectory point P5a in the direction toward the approximation line Lb2 along the lane width direction by the driving region width W is set as a second boundary point.

[0052] Similarly, for a group of three trajectory points P6a, P7a, and P8a, an approximation line Lb3 of the trajectory points P6a, P7a, and P8a is calculated. The area boundary setting unit 32b sets a point P6b, which is obtained by moving the trajectory point P6a by the driving area width W along the lane width direction in the direction from the trajectory point P6a toward the approximation line Lb3, as a second boundary point.

[0053] A point P7b obtained by moving the trajectory point P7a from the trajectory point P7a in the direction toward the approximation line Lb3 along the lane width direction by the driving region width W is set as a second boundary point. A point P8b obtained by moving the trajectory point P8a from the trajectory point P8a in the direction toward the approximate line Lb3 along the lane width direction by the driving region width W is set as a second boundary point.

[0054] In addition, the trajectory points P1a to P8a on the first target driving trajectory T1 and the points P1b to P8b obtained by moving the trajectory points P1a to P8a by the driving area width W as described above may be outside the driving lane of the host vehicle 1. 7(c) shows an example in which the trajectory point P3a is outside the driving lane of the vehicle 1. In such a case, the area boundary setting unit 32b may set a first boundary point P3c at the intersection (triangular plot) of a straight line Lw that passes through the trajectory point P3a and extends in the lane width direction, and the lane boundary line. Similarly, if any of points P1b to P8b obtained by moving the trajectory points P1a to P8a by the driving area width W is outside the driving lane of the vehicle 1, a second boundary point may be set at the intersection of a straight line Lw extending in the lane width direction through this point and the lane boundary line.

[0055] The area boundary setting unit 32b sets one of the first boundary points and the second boundary point set by moving the first boundary point, which is located further to the right, as the right boundary point that forms the right boundary line BR of the driving area Rt, and sets the other point located further to the left as the left boundary point that forms the left boundary line BL of the driving area Rt. The area boundary setting unit 32b sets a right boundary line BR by connecting the right boundary points, and sets a left boundary line BL by connecting the left boundary points, thereby setting a traveling area Rt as shown in FIG. 3(b).

[0056] 5, the second target traveling trajectory generating unit 33 generates a second target traveling trajectory T2 within the traveling area Rt set by the traveling area setting unit 32. For example, the second target running trajectory generating unit 33 may generate a second target running trajectory T2 in which at least one of the curvature and curvature change of the second target running trajectory T2 satisfies a predetermined constraint condition, or in which both of the curvature and curvature change satisfy predetermined constraint conditions.

[0057] The constraint condition regarding the curvature may be, for example, that an evaluation function including the curvature of the second target running trajectory T2 as a variable is a minimum value, or may be a value within a predetermined allowable range. Similarly, the constraint condition regarding the curvature change may be, for example, that an evaluation function including the curvature change of the second target running trajectory T2 as a variable is a minimum value, or may be a value within a predetermined allowable range. The second target running trajectory generating unit 33 may generate the second target running trajectory T2 that satisfies the constraint conditions, for example, by an optimization calculation method. For example, the second target running trajectory generating unit 33 may generate the second target running trajectory T2 by an optimization calculation method so that an evaluation function including the curvature and the curvature change of the second target running trajectory T2 as variables is minimized. Furthermore, for example, the second target traveling trajectory generating unit 33 may generate the second target traveling trajectory T2 that satisfies the constraint conditions by using a learning model such as a neural network, a random forest, or a support vector machine.

[0058] FIG. 8 is a flowchart of an example of a driving control method according to the second embodiment. In step S11, the lane boundary detection unit 30 detects the shape of the lane ahead of the host vehicle 1 in the driving lane in which the host vehicle 1 is traveling. In step S12, the first target running trajectory generating unit 31 generates a first target running trajectory T1.

[0059] In step S13, the area width setting unit 32a of the traveling area setting unit 32 sets the traveling area width W. In step S14, the area boundary setting unit 32b sets boundary points of the traveling area Rt based on the first target traveling trajectory T1 and the traveling area width W, thereby setting the traveling area Rt. In step S15, the second target running trajectory generating unit 33 generates the second target running trajectory T2. In step S16, the vehicle control unit 34 controls the host vehicle 1 to travel along the second target travel trajectory T2, after which the process ends.

[0060] Third embodiment The controller 18 of the third embodiment generates a second target running trajectory T2 to run within the running area Rt in sections where the first target running trajectory T1 generated by the first target running trajectory generating unit 31 is not smooth, and generates the second target running trajectory T2 to pass through points on the first target running trajectory T1 in sections where the first target running trajectory T1 is smooth. The controller 18 of the third embodiment has a similar configuration to the controller 18 of the first and second embodiments, and the same or similar components are given the same reference symbols and overlapping descriptions are omitted.

[0061] The travel area setting unit 32 generates an approximation line of a plurality of trajectory points on the first target travel trajectory T1, and sets a first section in which the distance from the trajectory points on the first target travel trajectory to the approximation line is equal to or greater than a threshold, and a second section in which the distance is less than the threshold. In the first section, the first target travel trajectory T1 has a relatively rough shape, and in the second section, the first target travel trajectory T1 has a relatively smooth shape. Please refer to Fig. 9. For example, the traveling area setting unit 32 sets a group of at least three trajectory points on the first target traveling trajectory T1, and generates an approximation straight line (dashed line) of at least three trajectory points included in the group.

[0062] In the example of FIG. 9, a group of three adjacent trajectory points P0, P1a, and P2a, a group of trajectory points P2a, P3a, and P4a, a group of trajectory points P4a, P5a, and P6a, and a group of trajectory points P6a, P7a, and P8a are set. For the group of trajectory points P0, P1a, and P2a, a straight line Lc1 is calculated connecting the first point P0 and the third point P2a at both ends, and a distance D1 from the second point P1a between the first point P0 and the third point P2a to the straight line Lc1 is calculated.

[0063] Similarly, for the group of trajectory points P2a, P3a, and P4a, a straight line Lc2 connecting the first point P2a and the third point P4a is calculated, and a distance D2 from the second point P3a to the straight line Lc2 is calculated. For the group of trajectory points P4a, P5a, and P6a, a straight line Lc3 connecting the first point P4a and the third point P6a is calculated, and a distance D3 from the second point P5a to the straight line Lc3 is calculated.

[0064] For the group of trajectory points P6a, P7a, and P8a, a straight line Lc4 connecting the first point P6a and the third point P8a is calculated, and a distance D4 from the second point P7a to the straight line Lc4 is calculated. The traveling area setting unit 32 sets the section Sa in which the distances D1 and D2 are equal to or greater than the thresholds as a first section, and sets the section Sb in which the distances D3 and D4 are less than the thresholds as a second section.

[0065] The second target running trajectory generating unit 33 generates the second target running trajectory T2 so as to run within the running area Rt in the first section Sa, and generates the second target running trajectory T2 so as to pass through points on the first target running trajectory T1 in the second section Sb. Note that the running area setting unit 32 may omit setting the running area Rt in the second section Sb.

[0066] (Effects of the embodiment) (1) The external sensor 11 detects the shape of a lane ahead of the host vehicle 1 in the driving lane in which the host vehicle 1 is traveling. The controller 18 generates a first target traveling trajectory T1 in the driving lane ahead of the host vehicle 1 based on a traveling trajectory on which the vehicle has traveled in the past, sets a traveling area Rt in which the host vehicle 1 is traveling in the driving lane ahead of the host vehicle 1 based on the first target traveling trajectory T1, generates a second target traveling trajectory T2 for traveling within the traveling area Rt so as to satisfy a preset constraint condition for at least one of the curvature and the curvature change of the second target traveling trajectory T2, and controls the host vehicle 1 to travel along the second target traveling trajectory T2. For example, the controller 18 may generate the first target driving trajectory T1 by inputting the lane shape of the driving lane into a learning model 31a that learns the lane shapes and driving trajectories of lanes on which the vehicle has previously driven manually.

[0067] In this way, by generating the second target driving trajectory T2 within the driving region Rt set based on the first target driving trajectory T1, the advantages of the first target driving trajectory T1 generated by the learning model 31a that has learned the driving trajectory of the vehicle that was driven manually in the past can be utilized for the second target driving trajectory T2. For example, the second target driving trajectory T2 that is close to the trajectory intended by the occupant can be generated. Moreover, by generating the second target running trajectory T2 so that at least one of the curvature and the curvature change of the second target running trajectory T2 satisfies a preset constraint condition, a smooth second target running trajectory T2 can be generated.

[0068] (2) The controller 18 may generate the second target traveling trajectory T2 so that an evaluation function including at least one of the curvature and the curvature change of the second target traveling trajectory T2 as a variable, as a constraint condition, is minimized or is less than a threshold value. This makes it possible to generate a smooth second target traveling trajectory T2.

[0069] (3) The controller 18 may set the driving area width W, which is the width of the driving area Rt in the lane width direction, based on the shape of the first target driving trajectory T1. For example, the controller 18 may generate an approximation line of a plurality of points on the first target running trajectory T1, and set the running region width W based on the distance from the points on the first target running trajectory T1 to the approximation line.

[0070] For example, the controller 18 may set multiple groups including three adjacent points on the first target driving trajectory T1, calculate the distance from a second point sandwiched between a first point and a third point among the three adjacent points to a straight line connecting the first point and the third point for each group, and set the driving area width W based on the maximum distance from the second point to the straight line connecting the first point and the third point. As a result, the width of the driving area Rt can be set wider when the first target driving trajectory T1 is not smooth, and as a result, the degree of freedom of the passing points of the second target driving trajectory T2 can be increased, thereby generating a smoother second target driving trajectory T2.

[0071] (4) The controller 18 may generate an approximation line of at least three points on the first target driving trajectory T1, and set the points obtained by moving each of the at least three points by the driving area width W along the lane width direction in each direction toward the approximation line as boundary points of the driving area Rt. For example, the controller 18 may set a point on the first target driving trajectory T1 as a first boundary point, which is one of the boundary points at both ends of the lane width direction of the driving area Rt, and may set points obtained by moving each of at least three points along the lane width direction by the driving area width W in each direction toward the approximation line, as a second boundary point, which is the other of the boundary points at both ends. This allows the travel area Rt to be set in a shape that has characteristics roughly similar to the shape of the first target travel trajectory T1. Also, the travel area width W of the travel area Rt can be adjusted according to the smoothness of the first target travel trajectory T1.

[0072] (5) When the controller 18 moves any of the at least three points in each direction from any of the three points toward the approximation line along the lane width direction by the driving area width W and the point moves outside the driving lane, the controller 18 may set a boundary point at the intersection of the line extending in the lane width direction through the moved point and the lane boundary line. In addition, when a point on the first target driving trajectory T1 is outside the driving lane, the controller 18 may set a first boundary point at the intersection of a straight line extending in the lane width direction through the point on the first target driving trajectory T1 that is outside the driving lane and the lane boundary line. This makes it possible to prevent the second target driving trajectory T2 from being generated so as to pass along the outside of the driving lane.

[0073] (6) The controller 18 may generate an approximation line of multiple points on the first target driving trajectory T1, set a first section in which the distance from the points on the first target driving trajectory T1 to the approximation line is equal to or greater than a threshold, and a second section in which the distance is less than the threshold, generate a second target driving trajectory T2 within the driving area Rt in the first section, and generate a second target driving trajectory T2 that passes through points on the first target driving trajectory T1 in the second section. As a result, in a section where the first target running trajectory T1 is smooth, a trajectory closer to the first target running trajectory T1 is generated as the second target running trajectory T2, so that the vehicle can be made to run along a trajectory that more closely matches the intention of the occupant. Also, in a section where the first target running trajectory T1 is smooth, the calculation of the running region Rt can be omitted, so that the calculation load of the controller 18 can be reduced.

[0074] (7) The learning model 31a may be a neural network that has learned the lane shape, driving trajectory, vehicle speed, and lateral and longitudinal acceleration of the vehicle of lanes on which the vehicle has been manually driven in the past. The controller 18 may generate the first target driving trajectory T1 by inputting the lane shape of the driving lane, and the speed, lateral and longitudinal acceleration of the vehicle 1 into the neural network. This makes it possible to predict the driving trajectory intended by the occupant and generate it as the first target driving trajectory T1. [Explanation of symbols]

[0075] 1... host vehicle, 10... driving control device, 11... external sensor, 12... vehicle sensor, 13... positioning device, 14... map database, 15... navigation device, 17... actuator, 18... controller, 18a... processor, 18b... storage device, 30... lane boundary detection unit, 31... first target driving trajectory generation unit, 31a... learning model, 31b... predicted trajectory calculation unit, 32... driving area setting unit, 32a... area width setting unit, 32b... area boundary setting unit, 33... second target driving trajectory generation unit, 34... vehicle control unit

Claims

1. Detecting a lane shape ahead of the host vehicle in a lane in which the host vehicle is traveling; generating a first target driving trajectory in the driving lane ahead of the host vehicle based on a driving locus on which the vehicle has traveled in the past; A driving area in which the host vehicle travels in the driving lane ahead of the host vehicle is set based on the first target driving trajectory; generating a second target running path for running within the running area so as to satisfy a preset constraint condition for at least one of a curvature and a curvature change of the second target running path; controlling the host vehicle so as to travel along the second target travel path; A driving control method comprising:

2. The driving control method according to claim 1, characterized in that the second target driving trajectory is generated so that an evaluation function including at least one of a curvature and a curvature change of the second target driving trajectory as a variable, as the constraint condition, is minimized or is less than a threshold value.

3. 2. The driving control method according to claim 1, further comprising the step of: setting a driving area width, which is a width of the driving area in a lane width direction, based on a shape of the first target driving trajectory.

4. Generate an approximation straight line of a plurality of points on the first target running trajectory; setting the travel area width based on a distance from a point on the first target travel trajectory to the approximation straight line; 4. The method of claim 3, wherein the vehicle speed is controlled by the vehicle speed sensor.

5. Setting a plurality of groups each including three adjacent points on the first target running trajectory; calculating, for each group, a distance from a second point between the first point and the third point among the three adjacent points to a straight line connecting the first point and the third point; setting the travel area width based on the maximum value of the distance from the second point to a straight line connecting the first point and the third point; 5. The method of claim 4, wherein the vehicle speed is controlled by the vehicle speed sensor.

6. Generate an approximation line of at least three points on the first target running trajectory; A point obtained by moving each of the at least three points in a direction toward the approximation line from the at least three points by the width of the driving area along the lane width direction is set as a boundary point of the driving area.

4. The method of claim 3, wherein the vehicle speed is controlled by the vehicle speed sensor.

7. A point on the first target driving trajectory is set as a first boundary point, which is one of boundary points at both ends of the driving area in the lane width direction; As a second boundary point, which is the other of the boundary points at both ends, a point is set by moving each of the at least three points in each direction toward the approximation line along the lane width direction by the driving area width.

7. The method of claim 6, wherein the vehicle speed is controlled by the vehicle speed sensor.

8. The driving control method described in claim 7, characterized in that when any of the at least three points is moved by the driving area width along the lane width direction in each direction from any of the at least three points toward the approximation line, and the point moves outside the driving lane, the boundary point is set at the intersection of a straight line extending in the lane width direction through the moved point and the lane boundary line.

9. The driving control method described in claim 7, characterized in that, when a point on the first target driving trajectory is outside the driving lane, the first boundary point is set at the intersection of a straight line extending in the lane width direction through the point on the first target driving trajectory that is outside the driving lane and a lane boundary line.

10. Generate an approximation straight line of a plurality of points on the first target running trajectory; a first section in which a distance from a point on the first target traveling trajectory to the approximation straight line is equal to or greater than a threshold, and a second section in which the distance is less than the threshold; The second target running trajectory is generated within the running area in the first section; In the second section, the second target running trajectory is generated, the second target running trajectory passing through a point on the first target running trajectory.

2. The method of claim 1, wherein the vehicle speed is controlled by the vehicle speed sensor.

11. A sensor for detecting a shape of a lane ahead of the host vehicle in which the host vehicle is traveling; a controller that generates a first target driving trajectory in the driving lane ahead of the host vehicle based on a driving trajectory on which the vehicle has traveled in the past, sets a driving area in which the host vehicle will travel in the driving lane ahead of the host vehicle based on the first target driving trajectory, generates a second target driving trajectory for traveling within the driving area so as to satisfy a preset constraint condition for at least one of a curvature and a curvature change of the second target driving trajectory, and controls the host vehicle to travel along the second target driving trajectory; A driving control device comprising:

Citation Information

Patent Citations

  • Vehicle travel control device, vehicle travel control system, and vehicle travel control method

    JP2018022353A