WO2022228358A1 - 一种自动驾驶避障方法及系统、存储介质 - Google Patents

一种自动驾驶避障方法及系统、存储介质 Download PDF

Info

Publication number
WO2022228358A1
WO2022228358A1 PCT/CN2022/088877 CN2022088877W WO2022228358A1 WO 2022228358 A1 WO2022228358 A1 WO 2022228358A1 CN 2022088877 W CN2022088877 W CN 2022088877W WO 2022228358 A1 WO2022228358 A1 WO 2022228358A1
Authority
WO
WIPO (PCT)
Prior art keywords
obstacle
vehicle
point
force
passable area
Prior art date
Application number
PCT/CN2022/088877
Other languages
English (en)
French (fr)
Inventor
修彩靖
梁伟强
郭继舜
Original Assignee
广州汽车集团股份有限公司
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by 广州汽车集团股份有限公司 filed Critical 广州汽车集团股份有限公司
Priority to US18/013,064 priority Critical patent/US20230303120A1/en
Publication of WO2022228358A1 publication Critical patent/WO2022228358A1/zh

Links

Images

Classifications

    • B—PERFORMING OPERATIONS; TRANSPORTING
    • B62—LAND VEHICLES FOR TRAVELLING OTHERWISE THAN ON RAILS
    • B62D—MOTOR VEHICLES; TRAILERS
    • B62D15/00—Steering not otherwise provided for
    • B62D15/02—Steering position indicators ; Steering position determination; Steering aids
    • B62D15/025—Active steering aids, e.g. helping the driver by actively influencing the steering system after environment evaluation
    • B62D15/0265—Automatic obstacle avoidance by steering
    • B—PERFORMING OPERATIONS; TRANSPORTING
    • B60—VEHICLES IN GENERAL
    • B60W—CONJOINT CONTROL OF VEHICLE SUB-UNITS OF DIFFERENT TYPE OR DIFFERENT FUNCTION; CONTROL SYSTEMS SPECIALLY ADAPTED FOR HYBRID VEHICLES; ROAD VEHICLE DRIVE CONTROL SYSTEMS FOR PURPOSES NOT RELATED TO THE CONTROL OF A PARTICULAR SUB-UNIT
    • B60W60/00—Drive control systems specially adapted for autonomous road vehicles
    • B60W60/001—Planning or execution of driving tasks
    • B60W60/0015—Planning or execution of driving tasks specially adapted for safety
    • B—PERFORMING OPERATIONS; TRANSPORTING
    • B60—VEHICLES IN GENERAL
    • B60W—CONJOINT CONTROL OF VEHICLE SUB-UNITS OF DIFFERENT TYPE OR DIFFERENT FUNCTION; CONTROL SYSTEMS SPECIALLY ADAPTED FOR HYBRID VEHICLES; ROAD VEHICLE DRIVE CONTROL SYSTEMS FOR PURPOSES NOT RELATED TO THE CONTROL OF A PARTICULAR SUB-UNIT
    • B60W60/00—Drive control systems specially adapted for autonomous road vehicles
    • B60W60/001—Planning or execution of driving tasks
    • B60W60/0011—Planning or execution of driving tasks involving control alternatives for a single driving scenario, e.g. planning several paths to avoid obstacles
    • B—PERFORMING OPERATIONS; TRANSPORTING
    • B60—VEHICLES IN GENERAL
    • B60W—CONJOINT CONTROL OF VEHICLE SUB-UNITS OF DIFFERENT TYPE OR DIFFERENT FUNCTION; CONTROL SYSTEMS SPECIALLY ADAPTED FOR HYBRID VEHICLES; ROAD VEHICLE DRIVE CONTROL SYSTEMS FOR PURPOSES NOT RELATED TO THE CONTROL OF A PARTICULAR SUB-UNIT
    • B60W30/00—Purposes of road vehicle drive control systems not related to the control of a particular sub-unit, e.g. of systems using conjoint control of vehicle sub-units
    • B60W30/08—Active safety systems predicting or avoiding probable or impending collision or attempting to minimise its consequences
    • B60W30/09—Taking automatic action to avoid collision, e.g. braking and steering
    • B—PERFORMING OPERATIONS; TRANSPORTING
    • B60—VEHICLES IN GENERAL
    • B60W—CONJOINT CONTROL OF VEHICLE SUB-UNITS OF DIFFERENT TYPE OR DIFFERENT FUNCTION; CONTROL SYSTEMS SPECIALLY ADAPTED FOR HYBRID VEHICLES; ROAD VEHICLE DRIVE CONTROL SYSTEMS FOR PURPOSES NOT RELATED TO THE CONTROL OF A PARTICULAR SUB-UNIT
    • B60W40/00—Estimation or calculation of non-directly measurable driving parameters for road vehicle drive control systems not related to the control of a particular sub unit, e.g. by using mathematical models
    • B60W40/02—Estimation or calculation of non-directly measurable driving parameters for road vehicle drive control systems not related to the control of a particular sub unit, e.g. by using mathematical models related to ambient conditions
    • B60W40/06—Road conditions
    • B60W40/072—Curvature of the road
    • B—PERFORMING OPERATIONS; TRANSPORTING
    • B60—VEHICLES IN GENERAL
    • B60W—CONJOINT CONTROL OF VEHICLE SUB-UNITS OF DIFFERENT TYPE OR DIFFERENT FUNCTION; CONTROL SYSTEMS SPECIALLY ADAPTED FOR HYBRID VEHICLES; ROAD VEHICLE DRIVE CONTROL SYSTEMS FOR PURPOSES NOT RELATED TO THE CONTROL OF A PARTICULAR SUB-UNIT
    • B60W60/00—Drive control systems specially adapted for autonomous road vehicles
    • B—PERFORMING OPERATIONS; TRANSPORTING
    • B62—LAND VEHICLES FOR TRAVELLING OTHERWISE THAN ON RAILS
    • B62D—MOTOR VEHICLES; TRAILERS
    • B62D15/00—Steering not otherwise provided for
    • B62D15/02—Steering position indicators ; Steering position determination; Steering aids
    • B62D15/021—Determination of steering angle
    • G—PHYSICS
    • G01—MEASURING; TESTING
    • G01C—MEASURING DISTANCES, LEVELS OR BEARINGS; SURVEYING; NAVIGATION; GYROSCOPIC INSTRUMENTS; PHOTOGRAMMETRY OR VIDEOGRAMMETRY
    • G01C21/00—Navigation; Navigational instruments not provided for in groups G01C1/00 - G01C19/00
    • G01C21/20—Instruments for performing navigational calculations
    • G—PHYSICS
    • G05—CONTROLLING; REGULATING
    • G05D—SYSTEMS FOR CONTROLLING OR REGULATING NON-ELECTRIC VARIABLES
    • G05D1/00—Control of position, course, altitude or attitude of land, water, air or space vehicles, e.g. using automatic pilots
    • G05D1/02—Control of position or course in two dimensions
    • B—PERFORMING OPERATIONS; TRANSPORTING
    • B60—VEHICLES IN GENERAL
    • B60W—CONJOINT CONTROL OF VEHICLE SUB-UNITS OF DIFFERENT TYPE OR DIFFERENT FUNCTION; CONTROL SYSTEMS SPECIALLY ADAPTED FOR HYBRID VEHICLES; ROAD VEHICLE DRIVE CONTROL SYSTEMS FOR PURPOSES NOT RELATED TO THE CONTROL OF A PARTICULAR SUB-UNIT
    • B60W2552/00—Input parameters relating to infrastructure
    • B60W2552/50—Barriers
    • B—PERFORMING OPERATIONS; TRANSPORTING
    • B60—VEHICLES IN GENERAL
    • B60W—CONJOINT CONTROL OF VEHICLE SUB-UNITS OF DIFFERENT TYPE OR DIFFERENT FUNCTION; CONTROL SYSTEMS SPECIALLY ADAPTED FOR HYBRID VEHICLES; ROAD VEHICLE DRIVE CONTROL SYSTEMS FOR PURPOSES NOT RELATED TO THE CONTROL OF A PARTICULAR SUB-UNIT
    • B60W2554/00—Input parameters relating to objects
    • B60W2554/40—Dynamic objects, e.g. animals, windblown objects
    • B60W2554/404—Characteristics
    • B60W2554/4049—Relationship among other objects, e.g. converging dynamic objects

Definitions

  • the present invention relates to the technical field of vehicles, in particular to an automatic driving obstacle avoidance method and system, and a computer-readable storage medium.
  • the artificial potential field method was originally proposed by Khatib in the 1980s based on the movement of electrons in an electric field. Specifically, it simulates the nature of the point charge electric field in nature, abstracting obstacles into charged particles, and the surrounding electric potential field. Intensity is negatively correlated with distance. Assuming that the controlled object has the same charge, according to the principle of the same-sex repulsion of electric bodies, the controlled object tends to be far away from the local maximum value of the obstacle to achieve obstacle avoidance. In the same way, the target has different charges, and according to the principle of the attraction of opposites in electric bodies, the controlled object is guided to approach this global minimum. Finally, the virtual potentials are accumulated to form a total potential field, and the negative gradient direction is the virtual force direction.
  • the basic idea is to abstract the poses of the controlled object, target and obstacle, quantify it into virtual potential or virtual force, and construct a unified virtual potential field or virtual force field in the environment where the controlled object is located, so as to plan the controlled object. exercise.
  • the artificial potential field method has attracted widespread attention due to its good real-time performance, high security, less environmental information required, and smoother planned paths.
  • the local minima causes "deadlock", forming a local minima, and it is impossible to continue to advance to the preview point; and, the kinematic constraints, dynamic constraints, and environmental constraints and artificial potential field constraints of autonomous vehicles are not considered.
  • the correlation makes the environmental adaptability and reliability of automatic driving obstacle avoidance unable to meet the driving safety requirements.
  • the purpose of the present invention is to propose an automatic driving obstacle avoidance method and system, and a computer-readable storage medium, so as to solve the defect that the artificial potential field method falls into a local minimum, as well as the environmental constraints of the unmanned vehicle in a structured urban environment and the vehicle itself the problem of constraints.
  • a first aspect of the present invention proposes an automatic driving obstacle avoidance method, which includes the following steps:
  • the obstacle information at least includes the coordinates of the obstacle
  • a resultant force is calculated according to the attractive force and the repulsive force, a steering wheel angle control command is generated according to the resultant force, and a vehicle actuator is controlled to execute the steering wheel angle control command to achieve obstacle avoidance.
  • the determining a passable area according to the obstacle information includes:
  • each obstacle has one or more visible boundaries
  • the deviation between the direction of the passable area and the current heading of the vehicle body, and the deviation between the direction of the passable area and the direction of the preview point select one of the passable areas to output as the optimal passable area.
  • the one or more passable areas are determined according to the visible boundary of the one or more obstacles, the center point of the vehicle body and the minimum turning radius of the vehicle, including:
  • two arcs tangent to the course of the vehicle are generated by the left and right boundary points of the obstacle and the point O, so that the left boundary point of the obstacle and the point O is a point on one of the arcs, the right boundary point of the obstacle and the point O are points on the other arc; wherein the area between the two arcs is an impassable area;
  • two turning radius arcs are generated according to the point O and the minimum turning radius of the vehicle, and according to the two turning radius arcs, it is determined whether the remaining area except the impassable area is a passable area, so as to obtain One or more passable areas; wherein the boundary of each passable area includes a left circular arc and a right circular arc.
  • obtaining the safety factor of the one or more passable areas according to the boundary of the passable area, the relative position of the obstacle and the vehicle including:
  • the reference arc and the left arc and the right arc intersect at points B and C respectively, and the two turning radius circles If the arcs intersect at points A and D respectively, the width coefficient of the passable area is calculated according to the length of the arc between the points B and C and the length of the arc between the points A and D;
  • obtaining the deviation between the direction of the passable area and the current heading of the vehicle body, and the deviation between the direction of the passable area and the direction of the preview point according to the boundary of the passable area specifically includes:
  • the deviation between the direction of the passable area and the direction of the preview point is calculated according to the arc radius of the connecting arc between the preview point and the center of mass of the vehicle, and the radius of the left arc and the radius of the right arc of the passable area.
  • calculating the attractive force of the preview point to the vehicle according to the preview point and a preset gravitational potential field function includes:
  • the longitudinal action parameters of gravity are calculated;
  • the coordinate difference and the lateral action distance of the gravitational field on the vehicle are calculated to obtain the lateral action parameters of the gravitational force;
  • the attractive force of the preview point to the vehicle is calculated according to the longitudinal action parameter of gravity, the lateral action parameter of gravity and a preset attractive force gain coefficient.
  • the calculation of the repulsion force of the obstacle to the vehicle according to the coordinates of the obstacles on both sides of the passable area and the preset repulsive force potential field function includes:
  • the longitudinal action parameters of the repulsion force are calculated; and the lateral coordinate difference between the obstacle and the center point of the vehicle body is calculated. value and the lateral action distance of the repulsion field to the vehicle to obtain the lateral action parameters of the repulsion force;
  • the repulsion force of the obstacle to the vehicle is calculated according to the longitudinal action parameter of the gravitational force, the transverse action parameter of the repulsion force, and a preset repulsion force gain coefficient.
  • calculating a resultant force according to the attractive force and the repulsive force, and generating a steering wheel angle control instruction according to the resultant force comprising:
  • the steering wheel angle is calculated according to the longitudinal and lateral components of the resultant force, and a steering wheel angle control instruction is generated according to the steering wheel angle.
  • a second aspect of the present invention provides an automatic driving obstacle avoidance system, including:
  • an obstacle information acquisition unit used to acquire obstacle information of the road; wherein the obstacle information at least includes the coordinates of the obstacle;
  • a passable area determination unit configured to determine a passable area according to the obstacle information
  • an attractive force calculation unit configured to determine a preview point according to the passable area, and calculate the attractive force of the preview point to the vehicle according to the preview point and a preset gravitational potential field function
  • a repulsion force calculation unit configured to calculate the repulsion force of the obstacle to the vehicle according to the coordinates of the obstacles on both sides of the passable area and a preset repulsion force potential field function
  • the obstacle avoidance control unit is configured to calculate a resultant force according to the attractive force and the repulsion force, generate a steering wheel angle control command according to the resultant force, and control the vehicle actuator to execute the steering wheel angle control command to achieve obstacle avoidance.
  • a third aspect of the present invention provides a computer-readable storage medium on which a computer program is stored, and when the computer program is executed by a processor, implements the steps of the method for automatic driving obstacle avoidance in the first aspect.
  • the invention proposes an automatic driving obstacle avoidance method and system, and a computer-readable storage medium, which are aimed at the defects of artificial potential field method falling into local minimum, and the problems of unmanned vehicle environment constraints and vehicle constraints in structured urban environment , determine the passable area according to the obstacle information of the road ahead of the unmanned vehicle, and then determine the preview point according to the passable area, and establish the gravitational potential field function according to the preview point.
  • the repulsion potential field function is established, and a new artificial potential field model is proposed.
  • the artificial potential field model can well reflect the potential field strength of the unmanned vehicle with environmental constraints in the virtual potential field, so as to solve the artificial potential field method. Falling into local minimal defects, as well as the environmental constraints of unmanned vehicles in structured urban environments and the constraints of the vehicle itself, improve the environmental adaptability and reliability of autonomous driving obstacle avoidance.
  • FIG. 1 is a flowchart of an obstacle avoidance method for automatic driving in an embodiment of the present invention.
  • FIG. 2 is a diagram illustrating an example of a visible boundary of an obstacle in an embodiment of the present invention.
  • FIG. 3 is a diagram illustrating an example of obstacle restraint in an embodiment of the present invention.
  • FIG. 4 is a schematic structural frame diagram of an automatic driving obstacle avoidance system in another embodiment of the present invention.
  • An embodiment of the present invention provides an automatic driving obstacle avoidance method, which is suitable for unmanned vehicles and realizes an obstacle avoidance function as an automatic driving system of unmanned vehicles.
  • the method according to the embodiment of the present invention includes the following: Steps S1 to S6:
  • Step S1 obtaining obstacle information of the road; wherein the obstacle information at least includes the coordinates of the obstacle;
  • the unmanned vehicle is equipped with obstacle sensing equipment.
  • the obstacle sensing equipment will detect the obstacle information of the road ahead in real time.
  • Other traffic objects that have an impact on car driving can be considered obstacles, such as vehicles, pedestrians, etc.; for most automatic driving obstacle avoidance systems, usually, obstacle information is sensed through sensing devices and outputs a series of The original data of the boundary coordinates of the obstacle;
  • Step S2 determining a passable area according to the obstacle information
  • Step S3 determining a preview point according to the passable area, and calculating the attractive force of the preview point to the vehicle according to the preview point and a preset gravitational potential field function;
  • the unmanned vehicle will dynamically determine the preview point during the driving process.
  • the preview point can be understood as a forward target point of the unmanned vehicle, that is, the unmanned vehicle is expected to pass through the passable area in the future.
  • the preview point should be a position point in the passable area, therefore, the preview point needs to be determined according to the passable area; usually, the preview point is a point on the center line of the road; The selection of the center line and the preview point is well known to those skilled in the field of automatic driving technology, and will not be repeated here.
  • Step S4 calculating the repulsion of the obstacle to the vehicle according to the coordinates of the obstacles on both sides of the passable area and the preset repulsive potential field function;
  • the unmanned vehicle when driving through the passable area, the unmanned vehicle will be affected by the obstacles on both sides of the passable area. Therefore, in the artificial potential field, the unmanned vehicle and the obstacle are repelled. , the repulsion between the two is related to the relative position between the two, so the repulsion potential field function can be established between the coordinates of the obstacle and the coordinates of the unmanned vehicle;
  • Step S5 Calculate the resultant force according to the attractive force and the repulsive force, generate a steering wheel angle control command according to the resultant force, and control the vehicle actuator to execute the steering wheel angle control command to avoid obstacles.
  • the method of this embodiment is aimed at the defects of the artificial potential field method falling into the local minimum and the problems of the environmental constraints and vehicle constraints of the unmanned vehicle in the structured urban environment, and the passable area is determined according to the obstacle information of the road ahead of the unmanned vehicle. , and then determine the preview point according to the passable area, and establish the gravitational potential field function according to the preview point. At the same time, according to the obstacles on both sides of the passable area, the repulsion potential field function is established, and a new artificial potential field model is proposed.
  • the artificial potential field model can well reflect the potential field strength of the unmanned vehicle with environmental constraints in the virtual potential field, so as to solve the defect of the artificial potential field method falling into the local minimum, and the unmanned vehicle in the structured urban environment.
  • the problem of environmental constraints and the constraints of the vehicle itself improves the environmental adaptability and reliability of autonomous driving obstacle avoidance.
  • the step S2 specifically includes steps S21 to S24:
  • Step S21 Determine the visible boundaries of one or more obstacles according to the obstacle information; each obstacle has one or more visible boundaries.
  • a series of coordinate points of the visible boundary of the obstacle are obtained through the sensor.
  • the visible boundary of the obstacle is a polygon representing the number of obstacles and has multiple vertices, An ellipse or a circle can be considered to be formed by an inscribed regular polygon.
  • the obstacle information measured in practice is based on the sensor as the coordinate origin. According to the installation position of the sensor on the vehicle, after coordinate transformation, the obstacle visible boundary information in the vehicle coordinate system is obtained. In the vehicle coordinate system, connect the line between the coordinate origin O and the vertices of the visible boundary of the obstacle to obtain N line segments, and a polygon that includes all these line segments and does not contain obstacles is on all sides of the polygon.
  • the edge of the obstacle in the middle is the boundary of the visible obstacle.
  • the four vertices of the obstacles O1 and O2 are connected with the origin O to obtain four line segments, which can include all the line segments and exclude them.
  • the polygon of the obstacle is O-A1-D1
  • the obstacle O2 is O-B2-A2-D2
  • the visible boundary of the obstacle O1 is A1-D1
  • the visible boundary of the obstacle O2 is A2-B2 and A2- D2.
  • Step S22 Determine one or more passable areas according to the visible boundary of the one or more obstacles, the center point O of the vehicle body and the minimum turning radius of the vehicle.
  • the step S22 may include:
  • two arcs tangent to the course of the vehicle are generated by the left and right boundary points of the obstacle and the point O, so that the left boundary point of the obstacle and the The point O is a point on one of the arcs, the right boundary point of the obstacle and the point O are points on the other arc; the area between the two arcs is impassable area; for example, if the midpoint F and point E of the obstacle O1 are the left and right boundary points, the arc OF and the arc OE can be obtained, and the area between the arc OF and the arc OE is an impassable area; the obstacle If the midpoint J and point H of the object O2 are the left and right boundary points, the arc OJ and the arc OH can be obtained, and the area between the arc OJ and the arc OH is an impassable area;
  • the generation principle of the two arcs tangent to the course of the vehicle is as follows: the center of the unmanned vehicle body and the boundary points on the left and right sides of the obstacle are used as points on the arc to form an envelope circle for the obstacle Arc, the arc is tangent to the direction of the vehicle, set at time t (O x , O y ) as the coordinates of the left boundary point or the right boundary point of the obstacle in the vehicle coordinate system, and the corresponding arc radius is calculated according to the following formula:
  • O Li the left side of the obstacle i Side arc curve segment
  • O Ri the right arc curve segment of obstacle i.
  • O Li (O Lix , O Liy ) are the coordinates of the left and right boundary points of the obstacle i obtained from the sensor, then O Li ,O Ri equation:
  • two turning radius arcs are generated, for example, the left and right two turning radius arcs of the point O in FIG. 3; and determined according to the two turning radius arcs Whether the remaining areas other than the impassable area are passable areas, so as to obtain one or more passable areas; wherein the boundary of each passable area includes a left circular arc and a right circular arc;
  • the existence of the minimum turning radius of the driverless vehicle is a constraint of the driverless vehicle itself.
  • the area that the driverless vehicle cannot reach is also the area that the driverless vehicle cannot pass, so the arc smaller than the minimum turning radius R min of the driverless vehicle cannot be executed. Therefore, in the arc obtained above, if
  • > R min , the impassable zone Z oneobstacle can be described by the equation of the left and right arcs as follows:
  • the areas other than the impassable areas are all passable areas. Generally speaking, there are many obstacles on the road. After the calculation of the above steps, multiple passable areas separated by the impassable areas will be obtained;
  • Step S23 obtain the safety factor of the one or more passable areas according to the boundary of the passable area, the relative position of the obstacle and the vehicle, and obtain the direction of the passable area and the current heading of the vehicle body according to the boundary of the passable area.
  • Deviation the deviation between the direction of the passable area and the direction of the preview point;
  • the area with higher safety should be selected, so the safety of the passable area S safety should be considered; and there are two main factors that affect the safety of the passable area, that is, the passable area
  • the width of the passable area and the relative position of the obstacles on both sides of the passable area and the vehicle, the closer the obstacle is to the vehicle, the greater the impact; the safety S safety of the i-th passable area can be expressed as the following formula:
  • Wi is the width coefficient of the ith passable area, indicating the width of the passable area
  • D i is the obstacle coefficient of the ith passable area, indicating the influence of the position of the obstacle
  • step S23 includes steps S231-S233:
  • Step S231 taking the point O as the center of the circle and the preset radius to generate a reference circular arc, the reference circular arc and the left circular arc and the right circular arc intersect at points B and C respectively, and the two If the turning radius arcs intersect at points A and D respectively, the width coefficient of the passable area is calculated according to the length of the arc between the points B and C and the length of the arc between the points A and D;
  • the origin of the vehicle coordinate system is used as the center of the circle, and the arc is made with any fixed radius.
  • the arc intersects the boundary arc of the passable area at points B and C, and is connected to the minimum turning radius of the driverless vehicle. If the arc intersects at points A and D, the width coefficient of the passable area that defines the ith passable area is:
  • Step S232 Determine the obstacle point I closest to the vehicle among the multiple obstacle points of the obstacles on both sides of the passable area, according to the length of the straight line between the point O and the point I, and the distance between the point B and the point C Calculate the obstacle coefficient from the length of the straight line, the straight line length between the point O and point C, and the width of the vehicle;
  • the obstacle coefficient of the ith passable area is:
  • W car is the width of the vehicle
  • OI, BC, and OC are the lengths of the straight line segment between two points
  • Step S233 calculating the safety factor of the passable area according to the width coefficient and the obstacle coefficient of the passable area;
  • S safety f(W i ,D i ) is an expression that expresses the relationship between safety and safety factor, width factor and obstacle factor.
  • the technical requirements are to adjust the weight values of the width coefficient and the obstacle coefficient. Because the parameters of the preset radius are different, the size of the obtained width coefficient will also be adjusted. Therefore, it can be appropriately adjusted according to the actual situation;
  • the selection of the passable area should also take into account the following two aspects: the deviation of the passable area related to comfort and the current heading of the vehicle body and the deviation of the passable area related to the driving efficiency and the direction of the preview point;
  • obtaining the deviation between the direction of the passable area and the current heading of the vehicle body, and the deviation between the direction of the passable area and the direction of the preview point according to the boundary of the passable area specifically includes:
  • G deviation (i) is the deviation between the direction of the passable area of the ith passable area and the direction of the preview point
  • R g is the difference between the preview point and the center of mass of the vehicle
  • Step S24 selecting one of the passable areas to output as the optimal passable area according to the safety factor, the deviation between the direction of the passable area and the current heading of the vehicle body, and the deviation between the direction of the passable area and the direction of the preview point;
  • the selection of a passable area can be calculated using the following cost function, which is the ith passable area cost function:
  • k s , k h , and kg are the weight coefficients of the safety factor, the heading deviation factor, and the target point deviation factor, respectively, and the corresponding weights can be set according to the weight of the three factors.
  • the passable area is determined by calculating the cost function of the passable area.
  • the minimum cost function value is the optimal passable area.
  • the maximum cost function threshold is also set.
  • the step S3 includes:
  • the longitudinal action parameters of gravity are calculated;
  • the coordinate difference and the lateral action distance of the gravitational field on the vehicle are calculated to obtain the lateral action parameters of the gravitational force;
  • the attractive force of the preview point to the vehicle is calculated according to the longitudinal action parameter of gravity, the lateral action parameter of gravity and a preset attractive force gain coefficient.
  • the application of artificial potential field to autonomous vehicle obstacle avoidance should meet four basic conditions. First, it must be able to avoid the inherent defects of traditional artificial potential field functions. Second, it also requires that the target point function of its potential field function reflects unmanned The dynamic target point of the driving vehicle, the potential field of the three potential field functions can reflect the impact strength of the structured urban environment on the driverless vehicle in all distance directions, and the fourth one of the variables has a certain relationship with the vehicle dynamics constraints. a mapping relationship. However, some characteristics of the GAUSSIAN function meet the conditions, so an artificial potential field based on the GAUSSIAN function is constructed.
  • the preset gravitational potential field function described in this embodiment is shown in the following formula:
  • (X, Y) is the coordinate of the center point O of the vehicle body in the global coordinate system
  • (X g , Y g ) is the coordinate of the preview point in the global coordinate system
  • k att is the attraction potential gain coefficient
  • d attx is the longitudinal action distance of the gravitational field on the vehicle
  • d atty is the lateral action distance of the gravitational field on the vehicle
  • I g is a preset integer that can reflect the boundary of the potential field strength.
  • the preview point is determined according to the passable area, and the attractive force of the preview point to the vehicle is calculated according to the preview point and the preset gravitational potential field function, as shown in the following formula. Show:
  • K att ′ is the attraction gain coefficient
  • K att ′ is the attraction gain coefficient
  • step S4 includes:
  • the longitudinal action parameters of the repulsion force are calculated; and the lateral coordinate difference between the obstacle and the center point of the vehicle body is calculated. value and the lateral action distance of the repulsion field to the vehicle to obtain the lateral action parameters of the repulsion force;
  • the repulsion force of the obstacle to the vehicle is calculated according to the longitudinal action parameter of the gravitational force, the transverse action parameter of the repulsion force, and a preset repulsion force gain coefficient.
  • the obstacle is detected by a perception system such as lidar, and after obtaining the obstacle boundary point cloud, the relative coordinates (longitudinal closest point) of the closest point of each obstacle to the vehicle and the relative coordinates of the most boundary point of each obstacle can be obtained. (lateral closest point).
  • the traditional repulsion potential model is the same as the repulsion potential at the obstacle boundary and the influence distance, and the repulsion potential increases sharply when approaching the obstacle, but for the unmanned vehicle, the influence distance of an obstacle to the unmanned vehicle And the repulsion potential is different.
  • the preset repulsive potential field function in this embodiment is shown in the following formula:
  • d repx (i) is the longitudinal action distance of the i-th obstacle repulsion field on the vehicle
  • d repy (i) is the lateral action distance of the i-th obstacle repulsion field on the vehicle
  • I o is a preset integer
  • (X, Y) are the coordinates of the center point O of the vehicle body in the global coordinate system
  • (X O (i), Y O (i)) and the i-th obstacle closest to the vehicle are at Coordinate in the global coordinate system
  • k rep is the repulsive potential gain coefficient.
  • the repulsion force of the obstacle to the vehicle is calculated according to the coordinates of the obstacles on both sides of the passable area and the preset repulsive force potential field function, as shown in the following formula:
  • k′ rep 4*k rep *I o , is the unit vector from the point O to the point of repulsion; is the longitudinal action parameter of the repulsion force, is the gravitational lateral action parameter.
  • step S5 specifically includes:
  • Step S51 calculate the angle between the vehicle and the preview point according to the coordinates of the preview point, and calculate the vertical and horizontal components of the attraction according to the angle and the attractive force, and according to the Calculate the longitudinal and lateral components of the repulsive force by calculating the included angle and the repulsive force;
  • Step S52 calculate the longitudinal component of the resultant force according to the longitudinal component of the attractive force and the longitudinal component of the repulsive force, and calculate the transverse component of the resultant force according to the transverse component of the attractive force and the transverse component of the repulsive force. force;
  • Step S53 Calculate the steering wheel angle according to the longitudinal and lateral components of the resultant force, and generate a steering wheel angle control instruction according to the steering wheel angle.
  • the direction of the resultant force determines the direction of movement of the controlled object.
  • the gravitational force and the repulsive force are decomposed into component forces on two coordinate axes respectively.
  • the ith repulsion point X ob (i) [x ob (i), y ob (i)], then the angle between the corresponding driverless car and the ith repulsion point :
  • ⁇ i arctan(y ob (i)/x ob (i))
  • the magnitude of the repulsive force on the horizontal and vertical coordinates is:
  • i s is the transmission ratio of the steering system.
  • another embodiment of the present invention provides an automatic driving obstacle avoidance system, which can be used to implement the method described in the above embodiment.
  • the system in this embodiment includes:
  • An obstacle information obtaining unit 1 configured to obtain obstacle information of the road; wherein the obstacle information at least includes the coordinates of the obstacle;
  • a passable area determination unit 2 configured to determine a passable area according to the obstacle information
  • Attraction calculation unit 3 configured to determine a preview point according to the passable area, and calculate the attractive force of the preview point to the vehicle according to the preview point and a preset gravitational potential field function;
  • a repulsive force calculation unit 4 configured to calculate the repulsive force of the obstacle to the vehicle according to the coordinates of the obstacles on both sides of the passable area and a preset repulsive force potential field function;
  • the obstacle avoidance control unit 5 is configured to calculate a resultant force according to the attractive force and the repulsion force, generate a steering wheel angle control command according to the resultant force, and control the vehicle actuator to execute the steering wheel angle control command to achieve obstacle avoidance.
  • the automatic driving obstacle avoidance system described in the above embodiment is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium.
  • Another embodiment of the present invention provides a computer-readable storage medium on which a computer program is stored, and when the computer program is executed by a processor, implements the steps of the automatic driving obstacle avoidance method described in the foregoing embodiment.
  • the computer-readable storage medium may include: any entity or device capable of carrying the computer program code, a recording medium, a USB flash drive, a removable hard disk, a magnetic disk, an optical disk, a computer memory, a read-only memory (ROM, Read-Only Memory), random access memory (RAM, Random Access Memory), electric carrier signal, telecommunication signal and software distribution medium, etc.

Landscapes

  • Engineering & Computer Science (AREA)
  • Mechanical Engineering (AREA)
  • Transportation (AREA)
  • Automation & Control Theory (AREA)
  • Radar, Positioning & Navigation (AREA)
  • Remote Sensing (AREA)
  • Human Computer Interaction (AREA)
  • Physics & Mathematics (AREA)
  • Chemical & Material Sciences (AREA)
  • Combustion & Propulsion (AREA)
  • General Physics & Mathematics (AREA)
  • Mathematical Physics (AREA)
  • Aviation & Aerospace Engineering (AREA)
  • Control Of Position, Course, Altitude, Or Attitude Of Moving Bodies (AREA)
  • Traffic Control Systems (AREA)

Abstract

一种自动驾驶避障方法及系统、存储介质，包括：获取道路的障碍物信息；其中障碍物信息至少包括障碍物的坐标；根据障碍物信息确定可通行区域；根据可通行区域确定预瞄点，并根据预瞄点和预设的引力势场函数计算预瞄点对本车的吸引力；根据可通行区域两侧的障碍物的坐标和预设的斥力势场函数计算障碍物对本车的斥力；根据吸引力和斥力计算合力；根据合力生成方向盘转角控制指令，并控制车辆执行机构执行方向盘转角控制指令以实现避障。该方法能够解决人工势场法陷入局部极小的缺陷，以及结构化城市环境下无人驾驶车环境约束及车辆自身约束的问题。

Description

一种自动驾驶避障方法及系统、存储介质
本申请要求于2021年4月25日提交中国专利局、申请号为202110448163.6、发明名称为“一种自动驾驶避障方法及系统、存储介质”的中国专利申请的优先权，上述专利的全部内容通过引用结合在本申请中。
技术领域
本发明涉及车辆技术领域，具体涉及一种自动驾驶避障方法及系统、计算机可读存储介质。
背景技术
人工势场法最初由Khatib在20世纪80年代根据电子在电场中的运动所提出的，具体来说，它模拟自然界中点电荷电场的性质，将障碍物抽象成带电粒子，其周边的电势场强度同距离负相关。设被控对象带同种电荷，根据电体同性相斥原理，被控对象趋向于远离障碍物这一局部最大值，实现避障。同理，目标物带异种电荷，根据电体异性相吸原理，引导被控对象趋向靠近这一全局极小值。最终，各项虚拟势累加构成总势场，其负梯度方向为虚拟力方向。其基本思想是将被控对象、目标物和障碍物的位姿抽象，量化为虚拟势或虚拟力，在被控对象所在环境中构建统一的虚拟势场或虚拟力场，从而规划被控对象的运动。
人工势场法由于实时性好，安全性比较高，其所需的环境信息较少，规划的路径比较平滑等优点而引起广泛关注，但目前人工势场法在自动驾驶的应用上，易陷入局部极小造成“死锁”，形成局部极小点，无法继续向预瞄点前进；以及，未考虑自动驾驶汽车运动学约束、动力学约束、及其所受的环境约束与人工势场的关联性，使得自动驾驶避障的环境适应性、可靠性无法满足驾驶安全要求。
发明内容
本发明的目的在于提出一种自动驾驶避障方法及系统、计算机可读存储介质，以解决人工势场法陷入局部极小的缺陷，以及结构化城市环境下无人驾驶车环境约束及车辆自身约束的问题。
为实现上述目的，本发明第一方面提出一种自动驾驶避障方法，包括如下步骤：
获取道路的障碍物信息；其中所述障碍物信息至少包括障碍物的坐标；
根据所述障碍物信息确定可通行区域；
根据所述可通行区域确定预瞄点，并根据所述预瞄点和预设的引力势场函数计算所述预瞄点对本车的吸引力；
根据所述可通行区域两侧的障碍物的坐标和预设的斥力势场函数计算障碍物对本车的斥力；
根据所述吸引力和所述斥力计算合力，根据所述合力生成方向盘转角控制指令，并控制车辆执行机构执行所述方向盘转角控制指令以实现避障。
优选地，所述根据所述障碍物信息确定可通行区域，包括：
根据所述障碍物信息确定一个或多个障碍物的可视边界；每一障碍物具有一条或多条可视边界；
根据所述一个或多个障碍物的可视边界、本车车体中心点O以及本车最小转弯半径确定一个或多个可通行区域；
根据可通行区域的边界、障碍物与本车的相对位置获取所述一个或多个可通行区域的安全系数，并根据可通行区域的边界获取可通行区域方向与车体当前航向的偏差、可通行区域方向与预瞄点方向的偏差；
根据所述安全系数、可通行区域方向与车体当前航向的偏差、可通行区域方向与预瞄点方向的偏差选择其中一个可通行区域输出作为最优可通行区域。
优选地，所述根据所述一个或多个障碍物的可视边界、本车车体中心点以及 本车最小转弯半径确定一个或多个可通行区域，包括：
对于任一障碍物，以该障碍物的左、右侧边界点和所述点O分别生成与本车航向相切的两个圆弧，使得该障碍物的左侧边界点和所述点O为其中一个圆弧上的点，该障碍物的右侧边界点和所述点O为其中另一个圆弧上的点；其中所述两个圆弧之间的区域为不可通行区域；
并且，根据所述点O以及本车最小转弯半径生成两个转弯半径圆弧，并根据所述两个转弯半径圆弧确定除所述不可通行区域以外的剩余区域是否为可通行区域，从而获得一个或多个可通行区域；其中每一可通行区域的边界包括左侧圆弧和右侧圆弧。
优选地，所述根据可通行区域的边界、障碍物与本车的相对位置获取所述一个或多个可通行区域的安全系数，包括：
以所述点O为圆心以及预设半径生成参考圆弧，所述参考圆弧与所述左侧圆弧和右侧圆弧分别相交于点B、点C，与所述两个转弯半径圆弧分别交于点A、点D，则根据所述点B、点C之间圆弧长度以及所述点A、点D两点之间圆弧长度计算可通行区域的宽度系数；
确定可通行区域两侧障碍物的多个障碍物点中距离本车最近的障碍物点I，根据所述点O、点I之间直线长度、所述点B、点C之间直线长度、所述点O、点C之间直线长度以及本车宽度计算障碍物系数；
根据可通行区域的宽度系数和障碍物系数计算可通行区域的安全系数。
优选地，所述根据可通行区域的边界获取可通行区域方向与车体当前航向的偏差、可通行区域方向与预瞄点方向的偏差，具体包括：
根据可通行区域的左侧圆弧半径、右侧圆弧半径计算可通行区域方向与车体当前航向的偏差；
并根据预瞄点与车辆质心的连接圆弧的圆弧半径以及可通行区域的左侧圆弧半径、右侧圆弧半径计算可通行区域方向与预瞄点方向的偏差。
优选地，所述根据所述预瞄点和预设的引力势场函数计算所述预瞄点对本车的吸引力，包括：
根据预瞄点与本车车体中心点之间的纵向坐标差值、引力场对本车的纵向作用距离计算得到引力纵向作用参数；并根据预瞄点与本车车体中心点之间的横向坐标差值、引力场对本车的横向作用距离计算得到引力横向作用参数；
根据所述引力纵向作用参数、所述引力横向作用参数以及预设的吸引力增益系数计算所述预瞄点对本车的吸引力。
优选地，所述根据所述可通行区域两侧的障碍物的坐标和预设的斥力势场函数计算障碍物对本车的斥力，包括：
根据障碍物与本车车体中心点之间的纵向坐标差值、斥力场对本车的纵向作用距离计算得到斥力纵向作用参数；并根据障碍物与本车车体中心点之间的横向坐标差值、斥力场对本车的横向作用距离计算得到斥力横向作用参数；
根据所述引力纵向作用参数、所述斥力横向作用参数以及预设的斥力增益系数计算所述障碍物对本车的斥力。
优选地，所述根据所述吸引力和所述斥力计算合力，根据所述合力生成方向盘转角控制指令，包括：
根据所述预瞄点的坐标计算本车与预瞄点之间的夹角，并根据所述夹角与所述吸引力计算吸引力的纵向分力和横向分力，并根据所述夹角与所述斥力计算斥力的纵向分力和横向分力；
根据所述吸引力的纵向分力和所述斥力的纵向分力计算合力的纵向分力，并根据所述吸引力的横向分力和所述斥力的横向分力计算合力的横向分力；
根据所述合力的纵向分力、横向分力计算方向盘转角，并根据该方向盘转角生成方向盘转角控制指令。
与上述方法对应，本发明第二方面提出一种自动驾驶避障系统，包括：
障碍物信息获取单元，用于获取道路的障碍物信息；其中所述障碍物信息至 少包括障碍物的坐标；
可通行区域确定单元，用于根据所述障碍物信息确定可通行区域；
吸引力计算单元，用于根据所述可通行区域确定预瞄点，并根据所述预瞄点和预设的引力势场函数计算所述预瞄点对本车的吸引力；
斥力计算单元，用于根据所述可通行区域两侧的障碍物的坐标和预设的斥力势场函数计算障碍物对本车的斥力；以及
避障控制单元，用于根据所述吸引力和所述斥力计算合力，根据所述合力生成方向盘转角控制指令，并控制车辆执行机构执行所述方向盘转角控制指令以实现避障。
本发明第三方面提出一种计算机可读存储介质，其上存储有计算机程序，所述计算机程序被处理器执行时实现第一方面所述自动驾驶避障方法的步骤。
本发明提出了一种自动驾驶避障方法及系统、计算机可读存储介质，其针对人工势场法陷入局部极小的缺陷，以及结构化城市环境下无人驾驶车环境约束及车辆约束的问题，根据无人驾驶车前行道路的障碍物信息确定可通行区域，然后根据该可通行区域确定预瞄点，并根据预瞄点建立引力势场函数，同时，根据可通行区域两侧障碍物建立斥力势场函数，从而提出了新的人工势场模型，该人工势场模型能够很好地反映具有环境约束的无人驾驶车在虚拟势场所受到的势场强度，从而解决人工势场法陷入局部极小的缺陷，以及结构化城市环境下无人驾驶车环境约束及车辆自身约束的问题，提高了自动驾驶避障的环境适应性、可靠性。
本发明的其它特征和优点将在随后的说明书中阐述。
附图说明
为了更清楚地说明本发明实施例或现有技术中的技术方案，下面将对实施例或现有技术描述中所需要使用的附图作简单地介绍，显而易见地，下面描述中的附图仅仅是本发明的一些实施例，对于本领域普通技术人员来讲，在不付出创造 性劳动的前提下，还可以根据这些附图获得其他的附图。
图1为本发明一实施例中一种自动驾驶避障方法的流程图。
图2为本发明一实施例中障碍物可视边界示例图。
图3为本发明一实施例中障碍物约束示例图。
图4为本发明另一实施例中一种自动驾驶避障系统的结构框架示意图。
具体实施方式
以下将参考附图详细说明本公开的各种示例性实施例、特征和方面。另外，为了更好的说明本发明，在下文的具体实施例中给出了众多的具体细节。本领域技术人员应当理解，没有某些具体细节，本发明同样可以实施。在一些实例中，对于本领域技术人员熟知的手段未作详细描述，以便于凸显本发明的主旨。
本发明一实施例提出一种自动驾驶避障方法，其适用于无人驾驶车，实现作为无人驾驶车的自动驾驶系统的一种避障功能，参阅图1，本发明实施例方法包括如下步骤S1～S6：
步骤S1、获取道路的障碍物信息；其中所述障碍物信息至少包括障碍物的坐标；
具体而言，无人驾驶车上配置有障碍物传感设备，无人驾驶车在道路上行驶时，障碍物传感设备会实时检测前方道路的障碍物信息，在道路中存在对无人驾驶车行驶具有影响的其他交通物体都可被认为是障碍物，例如是车辆、行人等；对于大多数自动驾驶避障系统而言，通常情况下，障碍物信息通过传感设备感知并输出一系列障碍物边界坐标点的原始数据；
步骤S2、根据所述障碍物信息确定可通行区域；
可以理解的是，根据所述障碍物信息可以确定前方道路那些区域是可通行区域，可通行区域的确定方式有很多种，在本实施例中不具体限定；
步骤S3、根据所述可通行区域确定预瞄点，并根据所述预瞄点和预设的引力势场函数计算所述预瞄点对本车的吸引力；
具体而言，无人驾驶车在行驶过程中，会动态地确定预瞄点，预瞄点可以理解为无人驾驶车的一个前进目标点，也即无人驾驶车预期通过可通行区域未来要到达的一个位置点，作为引导无人驾驶车行驶的点，也即在人工势场中预瞄点与无人驾驶车之间是吸引的，吸引无人驾驶车到达预瞄点；
可以理解的是，预瞄点应当是可通行区域中的位置点，因此，需要根据所述可通行区域确定预瞄点；通常情况下，预瞄点选取的是道路中心线上的点；道路中心线以及预瞄点的选取为自动驾驶技术领域的技术人员所熟知，此处不再赘述，本实施例中旨在利用预瞄点来计算引力势场的吸引力；
步骤S4、根据所述可通行区域两侧的障碍物的坐标和预设的斥力势场函数计算障碍物对本车的斥力；
具体而言，无人驾驶车在行驶通过所述可通行区域时，会受到所述可通行区域两侧的障碍物的影响，因此在人工势场中无人驾驶车与障碍物之间是排斥的，两者之间的排斥与两者之间的相对位置有关系，因此，可以建立障碍物坐标与无人驾驶车坐标建立斥力势场函数；
步骤S5、根据所述吸引力和所述斥力计算合力，根据所述合力生成方向盘转角控制指令，并控制车辆执行机构执行所述方向盘转角控制指令以实现避障。
本实施例方法针对人工势场法陷入局部极小的缺陷，以及结构化城市环境下无人驾驶车环境约束及车辆约束的问题，根据无人驾驶车前行道路的障碍物信息确定可通行区域，然后根据该可通行区域确定预瞄点，并根据预瞄点建立引力势场函数，同时，根据可通行区域两侧障碍物建立斥力势场函数，从而提出了新的人工势场模型，该人工势场模型能够很好地反映具有环境约束的无人驾驶车在虚拟势场所受到的势场强度，从而解决人工势场法陷入局部极小的缺陷，以及结构化城市环境下无人驾驶车环境约束及车辆自身约束的问题，提高了自动驾驶避障的环境适应性、可靠性。
基于上述实施例方法，在一个例子中，所述步骤S2，具体包括步骤S21～S24：
步骤S21、根据所述障碍物信息确定一个或多个障碍物的可视边界；每一障碍物具有一条或多条可视边界。
具体而言，通过传感器获得障碍物可视边界的一系列坐标点，本实施例中，对于障碍物可视边界给出如下定义：假设障碍物为多边形代表障碍物个数且具有多个顶点，对于椭圆或者圆可认为由内切正多边形构成。在实际中所测得的障碍物信息是以传感器为坐标原点，根据传感器在车辆上的安装位置，经过坐标转换，即得到在车辆坐标系下的障碍物可视边界信息。在车辆坐标系下，分别连接坐标原点O和障碍物可视边界的各个顶点的连线，得到N条线段，将所有这些线段都包含在内且不包含障碍物的多边形，在多边形的所有边中为障碍物的边，即为可视障碍物边界，如图2所示，障碍物O1、O2的四个顶点与原点O连线得到四条线段，能将所有线段都包含在内且不包含障碍物的多边形为O-A1-D1，障碍物O2为O-B2-A2-D2，则障碍物O1的可视边界为A1-D1，障碍物O2的可视边界为A2-B2和A2-D2。
步骤S22、根据所述一个或多个障碍物的可视边界、本车车体中心点O以及本车最小转弯半径确定一个或多个可通行区域。
示例性地，所述步骤S22可以包括：
参阅图3，对于任一障碍物，以该障碍物的左、右侧边界点和所述点O分别生成与本车航向相切的两个圆弧，使得该障碍物的左侧边界点和所述点O为其中一个圆弧上的点，该障碍物的右侧边界点和所述点O为其中另一个圆弧上的点；其中所述两个圆弧之间的区域为不可通行区域；例如，障碍物O1中点F和点E为其左、右侧边界点，则可以获得圆弧OF、圆弧OE，圆弧OF、圆弧OE之间的区域为不可通行区域；障碍物O2中点J和点H为其左、右侧边界点，则可以获得圆弧OJ、圆弧OH，圆弧OJ、圆弧OH之间的区域为不可通行区域；
具体而言，所述与本车航向相切的两个圆弧的生成原理如下：以无人驾驶车车体中心和障碍物左右侧边界点作为圆弧上的点对障碍物作包络圆弧，圆弧与车 辆所在航向相切，设在t时刻(O x,O y)为障碍物左侧边界点或右侧边界点在车辆坐标系下的坐标，对应的圆弧半径根据以下计算公式：
Figure PCTCN2022088877-appb-000001
对于无人驾驶车视觉范围内任一个障碍物i，总会得到两条与车辆航向相切的圆弧O Li，O Ri构成障碍物i的不可通行区域，其中O Li为障碍物i的左侧圆弧曲线段，O Ri为障碍物i右侧圆弧曲线段。在车辆坐标系，O为车体中心点，O Li(O Lix,O Liy)，O Ri(O Rix,O Riy)为从传感器得到的障碍物i的左、右侧边界点坐标，则O Li,O Ri方程：
O Li:f oLi(x,y)＝(x-R Li) 2+y 2-R Li 2＝0
y＞0
|x|＜|O Lix|
O Ri:f oRi(x,y)＝(x-R Ri) 2+y 2-R Ri 2＝0
y＞0
|x|＜|O Rix|
并且，根据所述点O以及本车最小转弯半径R min生成两个转弯半径圆弧，例如图3中的点O的左右2个转弯半径圆弧；并根据所述两个转弯半径圆弧确定除所述不可通行区域以外的剩余区域是否为可通行区域，从而获得一个或多个可通行区域；其中每一可通行区域的边界包括左侧圆弧和右侧圆弧；
具体而言，无人驾驶车存在最小转弯半径是无人驾驶车本身的约束，所述最小转弯半径是指当转向盘转到极限位置，无人驾驶车以最低稳定车速转向行驶时，外侧转向轮的中心在支承平面上滚过的轨迹圆半径；其在很大程度上表征了无人驾驶车能够通过狭窄弯曲地带或绕过不可越过的障碍物的能力；转弯半径越小，无人驾驶车的机动性能越好。无人驾驶车无法达到的区域也为无人驾驶车不可通行区域，因此小于无人驾驶车最小转弯半径R min的圆弧同样无法执行，所以在以上所得到的圆弧中，如果|R|＜R min，则圆弧外均为不可通行区域。对于 |R|＞＝R min，则不可通行区域Z oneobstacle可以用左右圆弧的方程描述如下：
Figure PCTCN2022088877-appb-000002
可以理解的是，除不可通行区域以外的区域均为可通行区域，通常来说道路上存在很多个障碍物，经过上述步骤的运算，会得到被不可通行区域所隔开的多个可通行区域；
步骤S23、根据可通行区域的边界、障碍物与本车的相对位置获取所述一个或多个可通行区域的安全系数，并根据可通行区域的边界获取可通行区域方向与车体当前航向的偏差、可通行区域方向与预瞄点方向的偏差；
具体而言，在多个可通行区域中要选取安全性较高的区域，因此要考虑可通行区域的安全性S safety；而影响可通行区域安全性的因素主要有两个，即可通行区域的宽度情况和可通行区域两边障碍物与本车的相对位置，障碍物与本车越近则影响越大；其中第i个可通行区域的安全性S safety可以表示为以下公式：
S safety＝f(W i,D i)
其中，W i为第i个可通行区域的宽度系数，表示可通行区域的宽度情况；D i为第i个可通行区域的障碍物系数，表示障碍物的位置影响；
示例性地，所述步骤S23包括步骤S231～S233：
步骤S231、以所述点O为圆心以及预设半径生成参考圆弧，所述参考圆弧与所述左侧圆弧和右侧圆弧分别相交于点B、点C，与所述两个转弯半径圆弧分别交于点A、点D，则根据所述点B、点C之间圆弧长度以及所述点A、点D两点之间圆弧长度计算可通行区域的宽度系数；
具体而言，参阅图3，以车辆坐标系原点为圆心，以任意固定半径做圆弧，圆弧与可通行区域边界圆弧交于B，C两点，并与无人驾驶车最小转弯半径圆弧相交于A，D两点，则定义第i个可通行区域的可通行区域的宽度系数为：
Figure PCTCN2022088877-appb-000003
其中，
Figure PCTCN2022088877-appb-000004
为BC间圆弧长度，
Figure PCTCN2022088877-appb-000005
为AD间圆弧长度；
步骤S232、确定可通行区域两侧障碍物的多个障碍物点中距离本车最近的障碍物点I，根据所述点O、点I之间直线长度、所述点B、点C之间直线长度、所述点O、点C之间直线长度以及本车宽度计算障碍物系数；
具体而言，两边障碍物的相对位置的影响则通过比较获得最近的障碍物点；在图3中即为I点，则第i个可通行区域的障碍物系数为：
D i＝(OI*BC)/(OC*W car)
其中，W car为车辆宽度，OI、BC、OC均为两点间直线段长度；
步骤S233、根据可通行区域的宽度系数和障碍物系数计算可通行区域的安全系数；
具体而言，S safety＝f(W i,D i)是一个表达安全性和安全系数与宽度系数和障碍物系数相关的表达式，的计算公式可以是加权求和的方式，具体可以根据实际技术要求进行调整宽度系数和障碍物系数的权重值，因为所述预设半径的参数不同，则得出的宽度系数的大小也会有调整，因此，可以根据实际情况适当调整；
此外，可通行区域的选择还应考虑以下两个方面，与舒适性相关的可通行区域与车体的当前航向偏差及与行车效率相关的可通行区域与预瞄点方向的偏差；
示例性地，所述根据可通行区域的边界获取可通行区域方向与车体当前航向的偏差、可通行区域方向与预瞄点方向的偏差，具体包括：
根据可通行区域的左侧圆弧半径、右侧圆弧半径计算可通行区域方向与车体当前航向的偏差；具体地，根据公式
Figure PCTCN2022088877-appb-000006
计算可通行区域方向与车体当前航向的偏差；其中，H deviation(i)为第i个可通行区域的可通行区域方向与车体当前航向的偏差偏差，R Li为第i个可通行区域的左侧圆弧半径，R Ri为第i个可通行区域的右侧圆弧半径；
并根据预瞄点与车辆质心的连接圆弧的圆弧半径以及可通行区域的左侧圆 弧半径、右侧圆弧半径计算可通行区域方向与预瞄点方向的偏差；
具体地，根据公式
Figure PCTCN2022088877-appb-000007
计算可通行区域方向与预瞄点方向的偏差；其中，G deviation(i)为第i个可通行区域的可通行区域方向与预瞄点方向的偏差，R g为预瞄点与车辆质心的连接圆弧的圆弧半径。
步骤S24、根据所述安全系数、可通行区域方向与车体当前航向的偏差、可通行区域方向与预瞄点方向的偏差选择其中一个可通行区域输出作为最优可通行区域；
示例性地，对于可通行区域的选择可用如下代价函数来计算，为第i个可通行区域代价函数：
C i＝k sS safety(i)+k hH deviation(i)+k gG deviation(i)
其中k s,k h,k g分别为安全因子、航向偏差因子、目标点偏差因子的权值系数，可根据三个因子的偏重设置相应权值。通过计算可通行区域代价函数确定可通行区域，代价函数值最小的即为最优可通行区域，同时，也要设定最大代价函数阈值，当大于阈值时说明代价太大，不可通行。
基于上述实施例方法，在一个例子中，所述步骤S3，包括：
根据预瞄点与本车车体中心点之间的纵向坐标差值、引力场对本车的纵向作用距离计算得到引力纵向作用参数；并根据预瞄点与本车车体中心点之间的横向坐标差值、引力场对本车的横向作用距离计算得到引力横向作用参数；
根据所述引力纵向作用参数、所述引力横向作用参数以及预设的吸引力增益系数计算所述预瞄点对本车的吸引力。
具体而言，人工势场应用于无人驾驶车避障应满足四个基本条件，首先要能够避免传统人工势场函数的固有缺陷，其二还要求其势场函数的目标点函数体现无人驾驶车动态的目标点，其三势场函数的势场则能够反映结构化城市环境对无人驾驶车在各个距离方向上的影响强度，其四其某个变量与车辆动力学约束条件 有某种映射关系。而GAUSSIAN函数的某些特性满足条件，因此构建基于GAUSSIAN函数的人工势场。本实施例中所述预设的引力势场函数如下公式所示：
Figure PCTCN2022088877-appb-000008
其中，(X,Y)为本车车体中心点O的在全局坐标系中的坐标，(X g,Y g)为预瞄点在全局坐标系中的坐标，k att为吸引势增益系数，d attx为引力场对本车的纵向作用距离，d atty为引力场对本车的横向作用距离，I g为预先设定的整数，能够反映势场强度的边界。
在本实施例中，所述根据所述可通行区域确定预瞄点，并根据所述预瞄点和预设的引力势场函数计算所述预瞄点对本车的吸引力，具体如下公式所示：
Figure PCTCN2022088877-appb-000009
其中，
Figure PCTCN2022088877-appb-000010
为由点O指向吸引点的单位向量，K att′为吸引力增益系数；
Figure PCTCN2022088877-appb-000011
为所述引力纵向作用参数，
Figure PCTCN2022088877-appb-000012
为所述引力横向作用参数。
在一例子中，所述步骤S4，包括：
根据障碍物与本车车体中心点之间的纵向坐标差值、斥力场对本车的纵向作用距离计算得到斥力纵向作用参数；并根据障碍物与本车车体中心点之间的横向坐标差值、斥力场对本车的横向作用距离计算得到斥力横向作用参数；
根据所述引力纵向作用参数、所述斥力横向作用参数以及预设的斥力增益系数计算所述障碍物对本车的斥力。
其中，通过感知系统如激光雷达探测障碍物，得到障碍物边界点云后，可以 得到每个障碍物距离车辆的最近点的相对坐标(纵向最近点)以及每个障碍物最边界点的相对坐标(侧向最近点)。传统的斥力势模型在障碍物边界斥力势以及影响距离是一样大的，且在靠近障碍物时斥力势急剧增加，但是对于无人驾驶车来说，一个障碍物对于无人驾驶车的影响距离以及斥力势是不一样的，在无人驾驶车的前进方向即纵向斥力势的影响距离应该远大于障碍物侧向对无人驾驶车施加的斥力势，斥力势的变化也应该光滑，过于急剧会导致车辆的执行机构无法执行，且影响车辆的舒适性。因此，本实施例中的所述预设的斥力势场函数如下公式所示：
Figure PCTCN2022088877-appb-000013
其中，d repx(i)为第i个障碍物斥力场对本车的纵向作用距离，d repy(i)为第i个障碍物斥力场对本车的横向作用距离，I o为预先设定的整数，(X,Y)为本车车体中心点O的在全局坐标系中的坐标，(X O(i),Y O(i))和第i个障碍物距离本车最近的点的在全局坐标系中的坐标，k rep为斥力势增益系数。
在本实施例中，所述根据所述可通行区域两侧的障碍物的坐标和预设的斥力势场函数计算障碍物对本车的斥力，具体如下公式所示：
Figure PCTCN2022088877-appb-000014
其中，
Figure PCTCN2022088877-appb-000015
为第i个障碍物斥力场对本车的斥力，k′ rep＝4*k rep*I o，
Figure PCTCN2022088877-appb-000016
为由点O指向斥力点的单位向量；
Figure PCTCN2022088877-appb-000017
为所述斥力纵向作用参数，
Figure PCTCN2022088877-appb-000018
为所述引力横向作用参数。
在本实施例中，所述步骤S5具体包括：
步骤S51、根据所述预瞄点的坐标计算本车与预瞄点之间的夹角，并根据所述夹角与所述吸引力计算吸引力的纵向分力和横向分力，并根据所述夹角与所述斥力计算斥力的纵向分力和横向分力；
步骤S52、根据所述吸引力的纵向分力和所述斥力的纵向分力计算合力的纵向分力，并根据所述吸引力的横向分力和所述斥力的横向分力计算合力的横向分力；
步骤S53、根据所述合力的纵向分力、横向分力计算方向盘转角，并根据该方向盘转角生成方向盘转角控制指令。
具体而言，合力方向即决定了被控对象的运动方向，在车体坐标系下，将引力和斥力分别分解为两坐标轴上的分力，在以车体坐标系为坐标轴建立的引力点函数上，选取引力点X g＝[x g,y g]，则无人驾驶车与引力点(即预瞄点)之间的夹角：
α＝arctan(y g/x g)
引力在横、纵坐标上的分力大小为：
Figure PCTCN2022088877-appb-000019
在车体坐标系上，第i个斥力点X ob(i)＝[x ob(i),y ob(i)]，则对应的无人驾驶车与第i个斥力点之间的夹角：
β i＝arctan(y ob(i)/x ob(i))
斥力在横、纵坐标上的分力大小为：
Figure PCTCN2022088877-appb-000020
无人驾驶车与合力的夹角：
δ＝arctan((F att(y g)-F rep(y ob(i)))/(F att(x g)-F rep(x ob(i))))
其中，i＝0,1…代表不同障碍物，则无人驾驶车与合力的方向盘转角为：
δ sw＝δ*i s
其中，i s为转向系统传动比。
与上述实施例方法对应，本发明另一实施例提出一种自动驾驶避障系统，可以用于实现上述实施例所述的方法，参阅图4，本实施例系统包括：
障碍物信息获取单元1，用于获取道路的障碍物信息；其中所述障碍物信息至少包括障碍物的坐标；
可通行区域确定单元2，用于根据所述障碍物信息确定可通行区域；
吸引力计算单元3，用于根据所述可通行区域确定预瞄点，并根据所述预瞄点和预设的引力势场函数计算所述预瞄点对本车的吸引力；
斥力计算单元4，用于根据所述可通行区域两侧的障碍物的坐标和预设的斥力势场函数计算障碍物对本车的斥力；以及
避障控制单元5，用于根据所述吸引力和所述斥力计算合力，根据所述合力生成方向盘转角控制指令，并控制车辆执行机构执行所述方向盘转角控制指令以实现避障。
以上所描述的系统实施例仅仅是示意性的，其中所述作为分离部件说明的单元可以是或者也可以不是物理上分开的，作为单元显示的部件可以是或者也可以不是物理单元，即可以位于一个地方，或者也可以分布到多个网络单元上。可以根据实际的需要选择其中的部分或者全部模块来实现本实施例方案的目的。
需说明的是，上述实施例所述系统与上述实施例所述方法对应，因此，上述实施例所述系统未详述部分可以参阅上述实施例所述方法的内容得到，即上述实施例方法记载的具体步骤内容可以理解为本实施例系统的所能够实现的功能，此处不再赘述。
并且，上述实施例所述自动驾驶避障系统若以软件功能单元的形式实现并作为独立的产品销售或使用时，可以存储在一个计算机可读取存储介质中。
本发明另一实施例提出一种计算机可读存储介质，其上存储有计算机程序，所述计算机程序被处理器执行时实现上述实施例所述自动驾驶避障方法的步骤。
具体而言，所述计算机可读存储介质可以包括：能够携带所述计算机程序代码的任何实体或装置、记录介质、U盘、移动硬盘、磁碟、光盘、计算机存储器、只读存储器(ROM，Read-Only Memory)、随机存取存储器(RAM，Random Access Memory)、电载波信号、电信信号以及软件分发介质等。
以上已经描述了本发明的各实施例，上述说明是示例性的，并非穷尽性的，并且也不限于所披露的各实施例。在不偏离所说明的各实施例的范围和精神的情况下，对于本技术领域的普通技术人员来说许多修改和变更都是显而易见的。本文中所用术语的选择，旨在最好地解释各实施例的原理、实际应用或对市场中的技术改进，或者使本技术领域的其它普通技术人员能理解本文披露的各实施例。

Claims (10)

  1. 一种自动驾驶避障方法，其特征在于，包括：
    获取道路的障碍物信息；其中所述障碍物信息至少包括障碍物的坐标；
    根据所述障碍物信息确定可通行区域；
    根据所述可通行区域确定预瞄点，并根据所述预瞄点和预设的引力势场函数计算所述预瞄点对本车的吸引力；
    根据所述可通行区域两侧的障碍物的坐标和预设的斥力势场函数计算障碍物对本车的斥力；
    根据所述吸引力和所述斥力计算合力，根据所述合力生成方向盘转角控制指令，并控制车辆执行机构执行所述方向盘转角控制指令以实现避障。
  2. 根据权利要求1所述的自动驾驶避障方法，其特征在于，所述根据所述障碍物信息确定可通行区域，包括：
    根据所述障碍物信息确定一个或多个障碍物的可视边界；每一障碍物具有一条或多条可视边界；
    根据所述一个或多个障碍物的可视边界、本车车体中心点O以及本车最小转弯半径确定一个或多个可通行区域；
    根据可通行区域的边界、障碍物与本车的相对位置获取所述一个或多个可通行区域的安全系数，并根据可通行区域的边界获取可通行区域方向与车体当前航向的偏差、可通行区域方向与预瞄点方向的偏差；
    根据所述安全系数、可通行区域方向与车体当前航向的偏差、可通行区域方向与预瞄点方向的偏差选择其中一个可通行区域输出作为最优可通行区域。
  3. 根据权利要求2所述的自动驾驶避障方法，其特征在于，所述根据所述一个或多个障碍物的可视边界、本车车体中心点以及本车最小转弯半径确定一个或多个可通行区域，包括：
    对于任一障碍物，以该障碍物的左、右侧边界点和所述点O分别生成与本车 航向相切的两个圆弧，使得该障碍物的左侧边界点和所述点O为其中一个圆弧上的点，该障碍物的右侧边界点和所述点O为其中另一个圆弧上的点；其中所述两个圆弧之间的区域为不可通行区域；
    并且，根据所述点O以及本车最小转弯半径生成两个转弯半径圆弧，并根据所述两个转弯半径圆弧确定除所述不可通行区域以外的剩余区域是否为可通行区域，从而获得一个或多个可通行区域；其中每一可通行区域的边界包括左侧圆弧和右侧圆弧。
  4. 根据权利要求3所述的自动驾驶避障方法，其特征在于，所述根据可通行区域的边界、障碍物与本车的相对位置获取所述一个或多个可通行区域的安全系数，包括：
    以所述点O为圆心以及预设半径生成参考圆弧，所述参考圆弧与所述左侧圆弧和右侧圆弧分别相交于点B、点C，与所述两个转弯半径圆弧分别交于点A、点D，则根据所述点B、点C之间圆弧长度以及所述点A、点D两点之间圆弧长度计算可通行区域的宽度系数；
    确定可通行区域两侧障碍物的多个障碍物点中距离本车最近的障碍物点I，根据所述点O、点I之间直线长度、所述点B、点C之间直线长度、所述点O、点C之间直线长度以及本车宽度计算障碍物系数；
    根据可通行区域的宽度系数和障碍物系数计算可通行区域的安全系数。
  5. 根据权利要求2所述的自动驾驶避障方法，其特征在于，所述根据可通行区域的边界获取可通行区域方向与车体当前航向的偏差、可通行区域方向与预瞄点方向的偏差，具体包括：
    根据可通行区域的左侧圆弧半径、右侧圆弧半径计算可通行区域方向与车体当前航向的偏差；
    并根据预瞄点与车辆质心的连接圆弧的圆弧半径以及可通行区域的左侧圆弧半径、右侧圆弧半径计算可通行区域方向与预瞄点方向的偏差。
  6. 根据权利要求1所述的自动驾驶避障方法，其特征在于，所述根据所述预瞄点和预设的引力势场函数计算所述预瞄点对本车的吸引力，包括：
    根据预瞄点与本车车体中心点之间的纵向坐标差值、引力场对本车的纵向作用距离计算得到引力纵向作用参数；并根据预瞄点与本车车体中心点之间的横向坐标差值、引力场对本车的横向作用距离计算得到引力横向作用参数；
    根据所述引力纵向作用参数、所述引力横向作用参数以及预设的吸引力增益系数计算所述预瞄点对本车的吸引力。
  7. 根据权利要求1所述的自动驾驶避障方法，其特征在于，所述根据所述可通行区域两侧的障碍物的坐标和预设的斥力势场函数计算障碍物对本车的斥力，包括：
    根据障碍物与本车车体中心点之间的纵向坐标差值、斥力场对本车的纵向作用距离计算得到斥力纵向作用参数；并根据障碍物与本车车体中心点之间的横向坐标差值、斥力场对本车的横向作用距离计算得到斥力横向作用参数；
    根据所述引力纵向作用参数、所述斥力横向作用参数以及预设的斥力增益系数计算所述障碍物对本车的斥力。
  8. 根据权利要求1所述的自动驾驶避障方法，其特征在于，所述根据所述吸引力和所述斥力计算合力，根据所述合力生成方向盘转角控制指令，包括：
    根据所述预瞄点的坐标计算本车与预瞄点之间的夹角，并根据所述夹角与所述吸引力计算吸引力的纵向分力和横向分力，并根据所述夹角与所述斥力计算斥力的纵向分力和横向分力；
    根据所述吸引力的纵向分力和所述斥力的纵向分力计算合力的纵向分力，并根据所述吸引力的横向分力和所述斥力的横向分力计算合力的横向分力；
    根据所述合力的纵向分力、横向分力计算方向盘转角，并根据该方向盘转角生成方向盘转角控制指令。
  9. 一种自动驾驶避障系统，其特征在于，包括：
    障碍物信息获取单元，用于获取道路的障碍物信息；其中所述障碍物信息至少包括障碍物的坐标；
    可通行区域确定单元，用于根据所述障碍物信息确定可通行区域；
    吸引力计算单元，用于根据所述可通行区域确定预瞄点，并根据所述预瞄点和预设的引力势场函数计算所述预瞄点对本车的吸引力；
    斥力计算单元，用于根据所述可通行区域两侧的障碍物的坐标和预设的斥力势场函数计算障碍物对本车的斥力；
    避障控制单元，用于根据所述吸引力和所述斥力计算合力，根据所述合力生成方向盘转角控制指令，并控制车辆执行机构执行所述方向盘转角控制指令以实现避障。
  10. 一种计算机可读存储介质，其上存储有计算机程序，其特征在于，所述计算机程序被处理器执行时实现权利要求1～8中任一项所述自动驾驶避障方法的步骤。
PCT/CN2022/088877 2021-04-25 2022-04-25 一种自动驾驶避障方法及系统、存储介质 WO2022228358A1 (zh)

Priority Applications (1)

Application Number Priority Date Filing Date Title
US18/013,064 US20230303120A1 (en) 2021-04-25 2022-04-25 Method for obstacle-avoidance in self-driving, system, and storage medium

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
CN202110448163.6 2021-04-25
CN202110448163.6A CN115230729A (zh) 2021-04-25 2021-04-25 一种自动驾驶避障方法及系统、存储介质

Publications (1)

Publication Number Publication Date
WO2022228358A1 true WO2022228358A1 (zh) 2022-11-03

Family

ID=83666080

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2022/088877 WO2022228358A1 (zh) 2021-04-25 2022-04-25 一种自动驾驶避障方法及系统、存储介质

Country Status (3)

Country Link
US (1) US20230303120A1 (zh)
CN (1) CN115230729A (zh)
WO (1) WO2022228358A1 (zh)

Families Citing this family (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN116101327A (zh) * 2023-04-14 2023-05-12 北京集度科技有限公司 行驶路径规划方法、装置、车辆及存储介质
CN117093005B (zh) * 2023-10-16 2024-01-30 华东交通大学 一种智能汽车自主避障方法

Citations (9)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
AU4604593A (en) * 1989-12-11 1993-12-02 Caterpillar Inc. Integrated vehicle positioning and navigation system apparatus and method
CN101436300A (zh) * 2008-12-10 2009-05-20 东软集团股份有限公司 障碍物分割方法及装置
CN102160006A (zh) * 2008-07-15 2011-08-17 空中侦察辨识和避免技术有限责任公司 用于避免碰撞的系统和方法
CN105764773A (zh) * 2013-08-29 2016-07-13 罗伯特·博世有限公司 用于控制车辆的方法
CN108897216A (zh) * 2018-07-03 2018-11-27 吉林大学 一种基于区域虚拟力场的智能汽车模型预测决策方法
CN110244713A (zh) * 2019-05-22 2019-09-17 江苏大学 一种基于人工势场法的智能车辆换道轨迹规划系统及方法
CN111750866A (zh) * 2020-07-07 2020-10-09 吉林大学 基于区域虚拟力场的智能汽车横纵向耦合路径规划方法
CN113515125A (zh) * 2021-07-05 2021-10-19 中国石油大学(华东) 一种无人驾驶汽车全工况避障控制方法及性能评价方法
CN113805597A (zh) * 2021-09-28 2021-12-17 福州大学 基于粒子群算法的障碍物自我保护人工势场法局部路径规划方法

Patent Citations (9)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
AU4604593A (en) * 1989-12-11 1993-12-02 Caterpillar Inc. Integrated vehicle positioning and navigation system apparatus and method
CN102160006A (zh) * 2008-07-15 2011-08-17 空中侦察辨识和避免技术有限责任公司 用于避免碰撞的系统和方法
CN101436300A (zh) * 2008-12-10 2009-05-20 东软集团股份有限公司 障碍物分割方法及装置
CN105764773A (zh) * 2013-08-29 2016-07-13 罗伯特·博世有限公司 用于控制车辆的方法
CN108897216A (zh) * 2018-07-03 2018-11-27 吉林大学 一种基于区域虚拟力场的智能汽车模型预测决策方法
CN110244713A (zh) * 2019-05-22 2019-09-17 江苏大学 一种基于人工势场法的智能车辆换道轨迹规划系统及方法
CN111750866A (zh) * 2020-07-07 2020-10-09 吉林大学 基于区域虚拟力场的智能汽车横纵向耦合路径规划方法
CN113515125A (zh) * 2021-07-05 2021-10-19 中国石油大学(华东) 一种无人驾驶汽车全工况避障控制方法及性能评价方法
CN113805597A (zh) * 2021-09-28 2021-12-17 福州大学 基于粒子群算法的障碍物自我保护人工势场法局部路径规划方法

Also Published As

Publication number Publication date
CN115230729A (zh) 2022-10-25
US20230303120A1 (en) 2023-09-28

Similar Documents

Publication Publication Date Title
US11970168B2 (en) Vehicle trajectory modification for following
CN109375632B (zh) 自动驾驶车辆实时轨迹规划方法
JP7480129B2 (ja) 圧力設定点を使用する車両サスペンションシステムの制御
CN112389427B (zh) 车辆轨迹优化方法、装置、电子设备和存储介质
CN112677995B (zh) 一种车辆轨迹规划方法、装置、存储介质及设备
US10919520B1 (en) Integrated chassis control
US11573571B2 (en) Trajectory generation and optimization using closed-form numerical integration in route-relative coordinates
WO2022228358A1 (zh) 一种自动驾驶避障方法及系统、存储介质
WO2021126651A1 (en) Prediction on top-down scenes based on object motion
CN114600053A (zh) 基于碰撞区域的行驶路线修改
JP2022527072A (ja) 属性に基づく歩行者の予測
US11409284B2 (en) Relaxation optimization model to plan an open space trajectory for autonomous vehicles
JP2022504979A (ja) 応答性のある車両制御
US11142188B2 (en) Action-based reference systems for vehicle control
KR20150038776A (ko) 인프라 센서를 이용한 자동주차 시스템
US20210197819A1 (en) Vehicle control to join route
US11300968B2 (en) Navigating congested environments with risk level sets
US11780464B2 (en) Autonomous vehicle trajectory generation using velocity-based steering limits
WO2023024914A1 (zh) 车辆避让方法、装置、计算机设备和存储介质
US20220371613A1 (en) Vehicle trajectory determination
CN116261540A (zh) 碰撞避免规划系统
Lombard et al. Lateral control of an unmaned car using GNSS positionning in the context of connected vehicles
US11591011B1 (en) Dynamic vehicle steering and/or suspension constraints
Jie et al. Research on trajectory planning model of lane changing based on improved artificial potential field method
US20240157944A1 (en) Reinforcement learning for autonomous lane change

Legal Events

Date Code Title Description
121 Ep: the epo has been informed by wipo that ep was designated in this application

Ref document number: 22794828

Country of ref document: EP

Kind code of ref document: A1

NENP Non-entry into the national phase

Ref country code: DE

122 Ep: pct application non-entry in european phase

Ref document number: 22794828

Country of ref document: EP

Kind code of ref document: A1