A path planning method for intelligent vehicles by controlling the direction of virtual force

Abstract

The present invention discloses a method for intelligent vehicle path planning for controlling the direction of virtual force. Aiming at the defects of the traditional artificial potential field method in the field of intelligent vehicle path planning, such as the target being unreachable and being easy to fall into a local minimum point, the artificial potential field is improved, including a virtual gravitational potential field of a target point, a virtual repulsive potential field of an obstacle, and a virtual repulsive potential field of a road boundary; wherein the virtual gravitational potential field of the target point is divided into two action areas, the virtual repulsive potential field of the obstacle is optimized into an elliptical action area, and a virtual line potential field between lanes is added to the virtual repulsive potential field of the road boundary. In the virtual force field of the improved artificial potential field, the method controls the virtual gravitational direction of the target point on the intelligent vehicle and the virtual repulsive direction of the obstacle on the intelligent vehicle, and controls the two in mutually perpendicular directions, avoids the virtual resultant force being 0 and causing the intelligent vehicle to fall into a local minimum point, ensures that the intelligent vehicle can smoothly avoid obstacles and reach the target point, and has a simple algorithm and good real-time performance.

Description

Technical Field

[0001] The present invention relates to the field of intelligent vehicle path planning, and in particular to an intelligent vehicle path planning method for controlling the direction of a virtual force. Background Art

[0002] The traditional artificial potential field algorithm has concise logic and simple model structure. By constructing the virtual gravitational potential field of the target point and the virtual repulsive potential field of the obstacle, the smart car moves forward in the direction of the virtual resultant force. The common obstacle vehicle potential field is a circular domain field. When the field meets the safety distance required to cover the obstacle vehicle in the lateral direction, it will increase the unnecessary safety distance of the smart car in the lateral direction; in the setting of the road boundary potential field, only the two boundaries of the road are considered, and the potential field of the virtual line between the lanes is not set, which has the problem of smart cars occupying two lanes; when the smart car, the obstacle and the target point are on the same straight line, it is easy to fall into the local minimum point. The fundamental reason is that the virtual repulsion of the smart car is opposite to the virtual gravitational force, and the magnitude is equal at a certain point, resulting in the virtual resultant force of the smart car being 0 and unable to move forward.

[0003] In response to the problems existing in the above-mentioned traditional artificial potential field in the path planning of smart cars, the obstacle potential field can be optimized into an elliptical action area to reduce the lateral safety distance of the smart car; the virtual line potential field between lanes can be added to the road boundary potential field to avoid the smart car from occupying two lanes; and for the local minimum problem, Li Yuhao et al. invented an artificial potential field path planning method based on fuzzy control of escape force (CN201911086388.0). When it is detected that the smart car is trapped in a local minimum point, this method helps the smart car escape from the local minimum point by increasing the escape force and fuzzy controlling the escape force, but it increases the overall complexity of the algorithm; Xu Jinli et al. invented a path planning method, device and equipment based on the improved artificial potential field method (CN202010690134.6). When it is detected that the smart car is trapped in a local minimum point, a simulated annealing algorithm with a variable step size is used to help the smart car escape from the local minimum point, but it increases the computational complexity and complexity of the algorithm. Summary of the invention

[0004] The technical problem to be solved by the present invention is to provide a smart car path planning method for controlling the direction of virtual force in view of the artificial potential field path planning problem mentioned in the background technology.

[0005] The present invention adopts the following technical solutions to solve the above technical problems:

[0006] A method for intelligent vehicle path planning for controlling the direction of virtual force comprises the following steps:

[0007] Step 1), obtain the global environmental information of the current road section, establish a coordinate system, define the direction parallel to the road as the X-axis, and the direction perpendicular to the road as the Y-axis. The dotted line between two lanes corresponds to the Y-axis coordinate of 0, and the intersection of the perpendicular line drawn from the starting point of the intelligent vehicle to the lane dotted line and the lane dotted line is the coordinate origin. The global environmental information includes the starting position of the intelligent vehicle, the positions of obstacles, and the position of the target point. The positions of the obstacles include the positions of dynamic obstacle vehicles and static obstacles.

[0008] Step 2), obtain the road information of the current road section according to the established coordinate system, and construct a virtual gravitational potential field of the target point, a virtual repulsive potential field of the obstacle, and a virtual repulsive potential field of the road boundary. Among them, the virtual gravitational potential field of the target point is used to guide the intelligent vehicle to move forward towards the target point; the virtual repulsive potential field of the obstacle is used to guide the intelligent vehicle to avoid obstacles during driving; the virtual repulsive potential field of the road boundary is used to restrict the intelligent vehicle from crossing the lane boundary and not occupying the double lanes for driving. The direction of the virtual repulsive force generated by the virtual repulsive potential field of the road boundary is perpendicular to the X-axis.

[0009] Step 3), according to the situation of the virtual repulsive force and virtual gravitational force received by the intelligent vehicle at the real-time position, execute the direction control strategy of the virtual gravitational force and virtual repulsive force until the intelligent vehicle enters the predetermined target position. Among them, the direction control strategy of the virtual gravitational force and virtual repulsive force is specifically as follows:

[0010] Step 3.1), the direction of the virtual gravitational force of the target point on the intelligent vehicle is controlled by the Euclidean distance Δd between the intelligent vehicle and the target point. The action area of the target point potential field is divided into two regions by a circle with a radius of d0. When Δd > d0, the direction of the virtual gravitational force of the target point on the intelligent vehicle is controlled to be along the positive X-axis; when Δd < d0, the direction of the virtual gravitational force of the target point on the intelligent vehicle is controlled to be the direction from the intelligent vehicle to the target point.

[0011] Step 3.2), the direction of the virtual repulsive force of the obstacle on the intelligent vehicle is controlled by the lane where the obstacle is located. When the obstacle is in the right lane, the direction of the virtual repulsive force of the obstacle on the intelligent vehicle is controlled to be along the positive Y-axis; when the obstacle is in the left lane, the direction of the virtual repulsive force of the obstacle on the intelligent vehicle is controlled to be along the negative Y-axis.

[0012] Step 3.3), the direction of the virtual gravitational force and the direction of the virtual repulsive force received by the intelligent vehicle are perpendicular to each other, and there is no component of the virtual repulsive force in the opposite direction of the virtual gravitational force.

[0013] As a further optimized solution of the intelligent vehicle path planning method for controlling the direction of virtual forces in the present invention, the virtual repulsive potential field of the obstacle in step 2)

[0014] The virtual gravitational potential field of the target point is The virtual gravitational potential field of the target point is divided into two regions by a circle with a radius of d0;

[0015] Virtual repulsion function for road boundaries

[0016] Among them, k rep Repulsion gain coefficient, k att is the gravitational gain coefficient, k edge is the virtual repulsion gain coefficient of the road boundary, represents the Euclidean distance between the smart car and the obstacle, represents the Euclidean distance between the smart car and the target point, ρ0 represents the safety distance, and The action range of the obstacle potential field is optimized into an ellipse, where |X-X0| is the absolute value of the difference between the horizontal coordinate of the smart car and the horizontal coordinate of the obstacle, |Y-Y0| is the absolute value of the difference between the vertical coordinate of the smart car and the vertical coordinate of the obstacle, A is the lateral safety distance to avoid collision between the smart car and the obstacle, B is the longitudinal safety distance for the smart car to cross the obstacle, d is the standard width of the lane on one side, w is the width of the car, and v is the speed of the smart car.

[0017] Compared with the prior art, the present invention adopts the above technical solution and has the following beneficial effects:

[0018] (1) Based on the traditional artificial potential field, the improved artificial potential field algorithm of the present invention optimizes the potential field action area of ​​the obstacle vehicle into an ellipse, reducing the unnecessary lateral movement distance of the smart car when crossing the obstacle vehicle; divides the target point potential field into two areas according to the distance, and controls the direction of the virtual gravity according to different areas to better pull the smart car to avoid obstacles; adds a virtual line potential field between lanes to prevent the smart car from occupying two lanes;

[0019] (2) Based on the improved artificial potential field method, the present invention controls the direction of the virtual repulsive force of the obstacle on the smart car and the direction of the virtual attractive force of the target point on the smart car in two directions perpendicular to each other, so as to avoid the virtual resultant force being 0, thereby fundamentally eliminating the problem of the smart car falling into a local minimum point;

[0020] (3) The present invention prevents the smart car from falling into a local minimum point by controlling the direction of the virtual force. The algorithm is simple and has good real-time performance. BRIEF DESCRIPTION OF THE DRAWINGS

[0021] Figure 1 A control flow chart of the intelligent vehicle path planning method for controlling the direction of virtual force of the present invention;

[0022] Figure 2 A schematic diagram of virtual forces acting on the smart car of the present invention at different positions;

[0023] Figure 3 A distribution diagram of the virtual repulsive potential field of the road boundary of the present invention;

[0024] Figure 4 is the coordinate distribution diagram of the road environment information of the present invention;

[0025] Figure 5 is the simulation result diagram of the path planning of the intelligent vehicle for controlling the virtual force direction of the present invention. Specific embodiments

[0026] The present invention will be further described in detail below with reference to the accompanying drawings:

[0027] As Figure 1 shown, the present invention discloses a path planning method for an intelligent vehicle for controlling the virtual force direction, including the following steps:

[0028] Step 1), obtain the global environment information of the current road section, establish a coordinate system, define the direction parallel to the road as the X-axis, the direction perpendicular to the road as the Y-axis, the dotted line between two lanes corresponds to the Y-axis 0 coordinate, and the intersection of the perpendicular line from the starting point of the intelligent vehicle to the lane dotted line and the lane dotted line is the coordinate origin; the global environment information includes the starting position of the intelligent vehicle, the position of obstacles, and the position of the target point, and the position of the obstacles includes the position of the dynamic obstacle vehicle and the position of the static obstacle;

[0029] Step 2), obtain the road information of the current road section according to the established coordinate system, construct the virtual gravitational potential field of the target point, the virtual repulsive potential field of the obstacle, and the virtual repulsive potential field of the road boundary, wherein the virtual gravitational potential field of the target point is used to guide the intelligent vehicle to move forward towards the target point; the virtual repulsive potential field of the obstacle is used to guide the intelligent vehicle to avoid obstacles during driving; the virtual repulsive potential field of the road boundary is used to limit the intelligent vehicle not to cross the lane boundary and also not to occupy the double-lane driving; the direction of the virtual repulsive force generated by the virtual repulsive potential field of the road boundary is perpendicular to the X-axis;

[0030] Step 3), according to the situation of the virtual repulsive force and virtual gravitational force received by the intelligent vehicle at the real-time position, execute the direction control strategy of the virtual gravitational force and virtual repulsive force until the intelligent vehicle enters the predetermined target position, wherein the direction control strategy of the virtual gravitational force and virtual repulsive force is specifically as follows:

[0031] Step 3.1), the virtual gravitational direction of the target point on the intelligent vehicle is controlled by the Euclidean distance Δd between the intelligent vehicle and the target point. The action area of the target point potential field is divided into two areas by a circle with a radius of d0. When Δd > d0, the virtual gravitational direction of the target point on the intelligent vehicle is controlled to be along the positive X-axis direction; when Δd < d0, the virtual gravitational direction of the target point on the intelligent vehicle is controlled to be the direction from the intelligent vehicle to the target point;

[0032] Step 3.2), the direction of the virtual repulsive force of the obstacle on the smart car is controlled by the lane where the obstacle is located. When the obstacle is in the right lane, the direction of the virtual repulsive force of the obstacle on the smart car is controlled to be along the positive direction of the Y axis; when the obstacle is in the left lane, the direction of the virtual repulsive force of the obstacle on the smart car is controlled to be along the negative direction of the Y axis;

[0033] Step 3.3), the direction of the virtual gravitational force and the direction of the virtual repulsive force on the smart car are perpendicular to each other, and the virtual repulsive force does not have a component opposite to the virtual gravitational force.

[0034] The virtual repulsive potential field of the obstacle described in step 2)

[0035]

[0036] The virtual gravitational potential field of the target point is The virtual gravitational potential field of the target point is divided into two regions by a circle with a radius of d0;

[0037] Virtual repulsion function for road boundaries

[0038] Among them, k rep Repulsion gain coefficient, k att is the gravitational gain coefficient, k edge is the virtual repulsion gain coefficient of the road boundary, represents the Euclidean distance between the smart car and the obstacle, represents the Euclidean distance between the smart car and the target point, ρ0 represents the safety distance, and The action range of the obstacle potential field is optimized into an ellipse, where |X-X0| is the absolute value of the difference between the horizontal coordinate of the smart car and the horizontal coordinate of the obstacle, |Y-Y0| is the absolute value of the difference between the vertical coordinate of the smart car and the vertical coordinate of the obstacle, A is the lateral safety distance to avoid collision between the smart car and the obstacle, B is the longitudinal safety distance for the smart car to cross the obstacle, d is the standard width of the lane on one side, w is the width of the car, and v is the speed of the smart car.

[0039] As the smart car moves forward, the forces on the smart car at different positions are as follows: Figure 2As shown in the figure, when the smart car is at the starting point, the smart car is not in the circular area with a radius of d0 of the target point, and is subject to the virtual gravitational force generated by the target point along the positive direction of the X axis; when the smart car touches the elliptical action area of ​​the obstacle (static and dynamic) on the right lane, it is subject to the virtual repulsive force along the positive direction of the Y axis, while the direction of the virtual gravitational force remains unchanged, and the smart car drives from the right lane to the left lane under the action of the virtual combined force; when the smart car touches the elliptical action area of ​​the obstacle (static and dynamic) in the left lane, it is subject to the virtual repulsive force along the negative direction of the Y axis, while the direction of the virtual gravitational force remains unchanged, and the smart car drives from the left lane to the right lane under the action of the virtual combined force; when the smart car enters the circular action area with a radius of d0 of the target point, the direction of the virtual gravitational force of the target point on the smart car changes from the smart car to the target point, pulling the smart car to the target point.

[0040] In the right lane, the road boundary potential field action area is as follows Figure 3 As shown in the figure, in the area [-d, -d+w / 2], the road boundary potential field generates a virtual repulsion along the positive direction of the Y axis, and the virtual repulsion is large, preventing the smart car from driving out of the lane; in the area [-d+w / 2, -w / 2], there is no potential field effect, and the virtual force is 0; in the area [-w / 2, 0], the road boundary potential field generates a virtual repulsion along the negative direction of the Y axis, and the virtual force is small, preventing the smart car from occupying both lanes at the same time, but when the smart car contacts the obstacle potential field action area and needs to change lanes, it can overcome the road boundary virtual repulsion to change lanes; the road boundary potential field action area of ​​the left lane is symmetrical with the action area of ​​the right lane about the X axis, and the direction of the virtual repulsion in the corresponding area is opposite to that of the right lane.

[0041] In order to verify the effectiveness and feasibility of the present invention, the method is used to perform path planning simulation in the Matlab software platform. The road environment is arranged as follows: Figure 4 As shown, the starting position coordinates of the smart car are set to [0,1.75], which is in the left lane; three static obstacles are set, and their coordinates are [10,1.75], [25,-1.5], and [90,1.75] respectively; two dynamic obstacle vehicles are set, the starting coordinates of obstacle vehicle No. 1 are [40,1.5], and it travels in the opposite direction of the smart car, and the starting coordinates of obstacle vehicle No. 2 are [45,-1.5], and it travels in the same direction as the smart car; the target point coordinates are set to [99,1.75], which is in the left lane; at the same time, the standard road width d = 3.5 meters, the car width w = 1.6 meters, the safety distance ρ0 = 10 meters, and in the elliptical action area of ​​the obstacle, A = 5 meters, B = 2 meters, and the radius of the circle dividing the target point area d0 = 5 meters.

[0042] The simulation results of the intelligent vehicle path planning for controlling the direction of virtual force proposed by the present invention are as follows: Figure 5As shown in the figure, when the obstacle moves forward and encounters a static obstacle at [10,1.75], the smart car, the obstacle, and the target point are in the same straight line. The smart car successfully avoids the first obstacle, and then successfully avoids the static obstacle at [25,-1.5]. When approaching the obstacle vehicle No. 1 coming from the opposite direction, the smart car smoothly drives from the left lane to the right lane, and then successfully avoids the obstacle vehicle No. 2 traveling in the same direction and the static obstacle at [90,1.75]. Finally, the smart car enters the circular area with a radius of d0, the direction of the virtual gravity changes, and the smart car successfully reaches the target point.

[0043] It will be understood by those skilled in the art that, unless otherwise defined, all terms (including technical and scientific terms) used herein have the same meaning as those generally understood by those skilled in the art in the art to which the present invention belongs. It should also be understood that terms such as those defined in common dictionaries should be understood to have meanings consistent with the meanings in the context of the prior art, and will not be interpreted with idealized or overly formal meanings unless defined as herein.

[0044] The specific implementation methods described above further illustrate the objectives, technical solutions and beneficial effects of the present invention in detail. It should be understood that the above description is only a specific implementation method of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present invention should be included in the scope of protection of the present invention.

Claims

1. A method for intelligent vehicle path planning for controlling the direction of virtual force, characterized in that: It includes the following steps: Step 1), obtain the global environmental information of the current road section, establish a coordinate system, define the direction parallel to the road as the X-axis, the direction perpendicular to the road as the Y-axis, the dotted line between two lanes corresponds to the Y-axis 0 coordinate, and the intersection of the perpendicular line from the starting point of the intelligent vehicle to the lane dotted line and the lane dotted line is the coordinate origin; the global environmental information includes the starting position of the intelligent vehicle, the positions of obstacles, and the position of the target point, and the positions of the obstacles include the positions of dynamic obstacle vehicles and static obstacles; Step 2), obtain the road information of the current road section according to the established coordinate system, construct a virtual gravitational potential field of the target point, a virtual repulsive potential field of the obstacle, and a virtual repulsive potential field of the road boundary. Among them, the virtual gravitational potential field of the target point is used to guide the intelligent vehicle to move forward towards the target point; the virtual repulsive potential field of the obstacle is used to guide the intelligent vehicle to avoid obstacles during driving; the virtual repulsive potential field of the road boundary is used to restrict the intelligent vehicle from crossing the lane boundary and not occupying the double lanes for driving; the direction of the virtual repulsive force generated by the virtual repulsive potential field of the road boundary is perpendicular to the X-axis; The obstacle virtual repulsive potential field The virtual gravitational potential field of the target point is The virtual gravitational potential field of the target point is divided into two regions by a circle with a radius of d0; Virtual repulsion function for road boundaries Among them, k rep Repulsion gain coefficient, k att is the gravitational gain coefficient, k edge is the virtual repulsion gain coefficient of the road boundary, represents the Euclidean distance between the smart car and the obstacle, represents the Euclidean distance between the smart car and the target point, ρ0 represents the safety distance, and The action range of the obstacle potential field is optimized into an ellipse, where |X-X0| is the absolute value of the difference between the horizontal coordinate of the smart car and the horizontal coordinate of the obstacle, |Y-Y0| is the absolute value of the difference between the vertical coordinate of the smart car and the vertical coordinate of the obstacle, A is the lateral safety distance to avoid collision between the smart car and the obstacle, B is the longitudinal safety distance for the smart car to cross the obstacle, d is the standard width of the lane on one side, w is the width of the car, and v is the speed of the smart car; Step 3), according to the situation of the virtual repulsive force and virtual gravitational force received by the intelligent vehicle at the real-time position, execute the direction control strategy of the virtual gravitational force and virtual repulsive force until the intelligent vehicle enters the predetermined target position. Among them, the direction control strategy of the virtual gravitational force and virtual repulsive force is specifically as follows: Step 3.1), the direction of the virtual gravitational force of the target point on the intelligent vehicle is controlled by the Euclidean distance Δd between the intelligent vehicle and the target point. The action area of the target point potential field is divided into two areas by a circle with a radius of d0. When Δd>d0, the direction of the virtual gravitational force of the target point on the intelligent vehicle is controlled to be along the positive X-axis; when Δd<d0, the direction of the virtual gravitational force of the target point on the intelligent vehicle is controlled to be the direction from the intelligent vehicle to the target point; Step 3.2), the direction of the virtual repulsive force of the obstacle on the intelligent vehicle is controlled by the lane where the obstacle is located. When the obstacle is in the right lane, the direction of the virtual repulsive force of the obstacle on the intelligent vehicle is controlled to be along the positive Y-axis; when the obstacle is in the left lane, the direction of the virtual repulsive force of the obstacle on the intelligent vehicle is controlled to be along the negative Y-axis; Step 3.3), the direction of the virtual gravitational force and the virtual repulsive force received by the intelligent vehicle are perpendicular to each other, and there is no component of the virtual repulsive force in the reverse direction of the virtual gravitational force.

Images