CN114323051B

CN114323051B - Intersection driving track planning method and device and electronic equipment

Info

Publication number: CN114323051B
Application number: CN202210014675.6A
Authority: CN
Inventors: 曹鹏飞; 唐天龙; 徐修信; 韩志华
Original assignee: Suzhou Zhitu Technology Co Ltd
Current assignee: Suzhou Zhitu Technology Co Ltd
Priority date: 2022-01-07
Filing date: 2022-01-07
Publication date: 2024-02-02
Anticipated expiration: 2042-01-07
Also published as: CN114323051A

Abstract

The invention provides a planning method, a planning device and electronic equipment for a road junction running track, wherein a current position is taken as an initial node, and a mixed A-type algorithm is used for determining an expansion node; if the appointed curve exists between the extended node and the current target point according to the priority order of the target point, determining an optimal running path; if no curve is specified, confirming that the distance between two nodes is smaller than a preset threshold according to the priority order, and determining the path through backtracking; otherwise, the expansion node is used as a new initial node, and the steps are repeated to determine the path; and optimizing the optimal running path by adopting a time elastic band algorithm to obtain the intersection running track. According to the method, the optimal running path of the vehicle at the intersection can be automatically searched by adopting a mixed A-type algorithm, and the searched optimal running path is optimized by adopting a time elastic band algorithm, so that the intersection running track of the vehicle capable of actually running is obtained, and the planning efficiency and accuracy of the intersection running track can be improved because a virtual road network is not required to be constructed manually.

Description

Intersection driving track planning method and device and electronic equipment

Technical Field

The invention relates to the technical field of automatic driving, in particular to a method and a device for planning a road junction driving track and electronic equipment.

Background

In the related art, a track planning method of an automatic driving vehicle at an intersection mainly comprises the steps of manually fitting a virtual lane line at the intersection, constructing a road network, and generating a virtual road center line as a virtual reference line for planning; specifically, path planning is generally performed on a virtual reference line, then speed planning is performed, and finally, a final track is obtained by combining the path and the speed. In the mode, in order to cover driving scenes as many as possible, enough possible virtual lane lines are needed to be drawn manually at the intersections to connect different intersection driving-in points and driving-out points, when the number of the road lanes is large and the traffic condition is complex, constructing the road network consumes a great deal of time and energy, and the road network is quite easy to make mistakes, so that the efficiency of planning the path of the intersection is lower and the accuracy is poor.

Disclosure of Invention

The invention aims to provide a method, a device and electronic equipment for planning a road junction running track so as to improve the efficiency and the accuracy of road junction track planning.

The invention provides a planning method of a road junction running track, which comprises the following steps: acquiring coordinates of a current position of a vehicle at an intersection and coordinates of at least one target point of the vehicle after the vehicle passes through the intersection; wherein each target point has its corresponding travel priority; different target points correspond to different lanes; the method comprises the steps of taking a current position as an initial node, adopting a mixed A-type algorithm based on the coordinates of the current position, and determining the coordinates of an expansion node adjacent to the initial node according to the kinematic characteristics of a vehicle; sequentially judging whether a designated curve exists between the extended node and the current target point according to the order of the running priority of at least one target point from high to low; wherein the designated curve is a Reeds-Shepp curve or a Dubin curve; if a specified curve exists, determining an optimal running path of the vehicle based on the current target point, the extension node and the specified curve; if no designated curve exists between the extended node and each target point, judging whether the distance between the extended node and the current target point is smaller than a preset threshold value according to the order of the running priority of at least one target point from high to low; if the distance between the extended node and the current target point is smaller than a preset threshold value, starting from the current target point, sequentially backtracking forward based on the extended node, and determining an optimal running path of the vehicle; if the distance between the extended node and each target point is not smaller than the preset threshold value, the extended node is taken as a new initial node, and the step of determining the coordinates of the extended node adjacent to the initial node according to the kinematic characteristics of the vehicle by adopting a mixed A algorithm is repeatedly executed until the optimal running path of the vehicle is determined; and (3) optimizing the optimal running path by adopting a time elastic band algorithm, and determining the intersection running track of the vehicle.

Further, the step of obtaining coordinates of a current position of the vehicle at the intersection and coordinates of at least one target point after the vehicle passes through the intersection includes: acquiring coordinates of a current position of a vehicle at an intersection and an initial path of the vehicle; the initial path is a path between a running start point and a running end point of a vehicle which is planned in advance; determining a target lane after a vehicle passes through an intersection based on the initial path, and determining the driving priority of the target lane as a first priority; if other lanes in the same direction as the target lane exist, sequentially determining the driving priority of the other lanes; the farther the distance between the other lanes and the target lane is, the lower the driving priority of the other lanes is; selecting target points from the target lane and other lanes respectively, and determining coordinates of at least one target point; the driving priority corresponding to each target point is the driving priority of the lane where the target point is located.

Further, the step of determining coordinates of an expansion node adjacent to the initial node according to the vehicle kinematic characteristic by using the hybrid a-algorithm based on the coordinates of the current position with the current position as the initial node includes: acquiring a preset expansion step length, front and rear wheelbases of a vehicle and an angle value of front wheel rotation of the vehicle during each expansion step; and determining the coordinates of the expansion nodes adjacent to the initial node according to the vehicle kinematic characteristics by adopting a mixed A-based algorithm according to the coordinates of the current position, the expansion step length, the front and rear wheelbase and the angle value.

Further, the step of sequentially judging whether a specified curve exists between the extended node and the current target point in an order of high to low traveling priority of at least one target point includes: if the intersection belongs to a scene allowing reversing, sequentially judging whether a Reeds-Shepp curve exists between the extension node and the current target point according to the order of the running priority of at least one target point from high to low; if the intersection does not belong to the scene allowing reversing, judging whether a Dubin curve exists between the extension node and the current target point according to the order of the running priority of at least one target point from high to low.

Further, the step of optimizing the optimal driving path by adopting a time elastic band algorithm and determining the crossing driving track of the vehicle comprises the following steps: acquiring an incomplete cost function, a speed cost function, an acceleration cost function, an obstacle collision-free cost function and a time cost function of a vehicle; weighting the incomplete cost function, the speed cost function, the acceleration cost function, the obstacle collision-free cost function and the time cost function based on preset weight coefficients to obtain an objective function; and solving an optimal solution of the objective function to optimize the optimal running path and determine the intersection running track of the vehicle.

Further, the objective function is determined by the following formula: objective function f=ω ₁ f _{non-holonomic} +ω ₂ f _v +ω ₃ f _a +ω ₄ f _obstacle +ω ₅ f _t The method comprises the steps of carrying out a first treatment on the surface of the Wherein f _{non-holonomic} Representing an incomplete cost function of the vehicle; f (f) _v Representing a speed cost function of the vehicle; f (f) _a Representing an acceleration cost function of the vehicle; f (f) _obstacle A collision-free cost function representing an obstacle of the vehicle; f (f) _t Representing a time cost function of the vehicle; ω1, ω2, ω3, ω4, and ω5 are weight coefficients corresponding to the respective cost functions, respectively.

Further, the non-integrity cost function f of the vehicle _{non-holonomic} Is determined by the following formula:

wherein,(x _i ，y _i ，β _i ) Coordinates of an ith expansion node; (x) _i+1 ，y _i+1 ，β _i+1 ) Is the coordinates of the i+1th expansion node.

Further, the speed cost function of the vehicle is determined by the following formula:

wherein,

representing the speed of the vehicle at the ith expansion node; (x) _i ，y _i ，β _i ) Coordinates of an ith expansion node; (x) _i+1 ，y _i+1 ，β _i+1 ) Coordinates of the (i+1) th expansion node; v _min A preset first speed threshold value; v _max A preset second speed threshold value; and v _min ＜v _max 。

Further, the acceleration cost function of the vehicle is determined by the following formula:

wherein,

representing acceleration of the vehicle at the i-th expansion node; v _i Representing the speed of the vehicle at the ith expansion node; v _i+1 Indicating vehicle Speed of the vehicle at the i+1th expansion node; deltaT _i Representing a time interval between an i-th expansion node and an i+1th expansion node; deltaT _i+1 Representing the time interval between the i+1th expansion node and the i+2th expansion node.

Further, the obstacle collision-free cost function of the vehicle is determined by the following formula:

wherein x is _i And y _i Coordinate values of the ith expansion node in the x-axis direction and the y-axis direction respectively;andcoordinate values of the jth obstacle in the x-axis direction and the y-axis direction in a preset range of the ith expansion node respectively; d, d _min Is a preset interval threshold value between the vehicle and the obstacle.

Further, the time cost function of the vehicle is determined by the following formula:wherein,

Δti represents the time interval between the i-th expansion node and the i+1th expansion node.

Further, the expansion node corresponding to i=0 is the current position of the vehicle at the intersection.

The invention provides a planning device for a road junction running track, which comprises: the acquisition module is used for acquiring the coordinates of the current position of the vehicle at the intersection and the coordinates of at least one target point after the vehicle passes through the intersection; wherein each target point has its corresponding travel priority; different target points correspond to different lanes; the first determining module is used for determining the coordinates of the expansion node adjacent to the initial node according to the vehicle kinematic characteristics by adopting a mixed A-type algorithm based on the coordinates of the current position by taking the current position as the initial node; the judging module is used for sequentially judging whether a specified curve exists between the extended node and the current target point according to the order of the running priority of at least one target point from high to low; wherein the designated curve is a Reeds-Shepp curve or a Dubin curve; a second determining module for determining an optimal travel path of the vehicle based on the current target point, the extension node, and the specified curve if the specified curve exists; a third determining module, configured to sequentially determine whether a distance between the extended node and the current target point is smaller than a preset threshold according to an order of high-to-low driving priority of at least one target point if a specified curve does not exist between the extended node and each target point; if the distance between the extended node and the current target point is smaller than a preset threshold value, starting from the current target point, sequentially backtracking forward based on the extended node, and determining an optimal running path of the vehicle; the repeated execution module is used for repeatedly executing the steps of adopting a mixed A algorithm to determine the coordinates of the extended nodes adjacent to the initial node according to the kinematic characteristics of the vehicle until the optimal running path of the vehicle is determined, wherein if the distance between the extended node and each target point is not smaller than a preset threshold value, the extended node is taken as a new initial node; and the optimization processing module is used for optimizing the optimal running path by adopting a time elastic band algorithm and determining the intersection running track of the vehicle.

The invention provides an electronic device, which comprises a processor and a memory, wherein the memory stores machine executable instructions which can be executed by the processor, and the processor executes the machine executable instructions to realize the method for planning the crossing driving track of any one of the above.

The machine-readable storage medium provided by the invention stores machine-executable instructions which, when being called and executed by a processor, cause the processor to implement the method for planning the intersection driving track of any one of the above.

The invention provides a planning method, a planning device and electronic equipment for a road junction running track, which are used for acquiring the coordinates of the current position of a vehicle at the road junction and the coordinates of at least one target point of the vehicle after the vehicle passes through the road junction; the method comprises the steps of taking a current position as an initial node, adopting a mixed A-type algorithm based on the coordinates of the current position, and determining the coordinates of an expansion node adjacent to the initial node according to the kinematic characteristics of a vehicle; and sequentially judging whether a specified curve exists between the extended node and the current target point according to the order of the running priority of at least one target point from high to low. If a specified curve exists, determining an optimal running path of the vehicle based on the current target point, the extension node and the specified curve; if no designated curve exists between the extended node and each target point, sequentially judging whether the distance between the extended node and the current target point is smaller than a preset threshold according to the order of the running priority of at least one target point from high to low, if the distance between the extended node and the current target point is smaller than the preset threshold, starting from the current target point, sequentially tracing back forward based on the extended node, and determining the optimal running path of the vehicle; if the distance between the extended node and each target point is not smaller than the preset threshold value, the extended node is taken as a new initial node, and the step of determining the coordinates of the extended node adjacent to the initial node according to the kinematic characteristics of the vehicle by adopting a mixed A algorithm is repeatedly executed until the optimal running path of the vehicle is determined; and (3) optimizing the optimal running path by adopting a time elastic band algorithm, and determining the intersection running track of the vehicle. According to the method, the optimal running path of the vehicle at the intersection can be automatically searched by adopting a mixed A-type algorithm, and the searched optimal running path is optimized by adopting a time elastic band algorithm, so that the intersection running track of the vehicle capable of actually running is obtained, and the planning efficiency and accuracy of the intersection running track can be improved because a virtual road network is not required to be constructed manually.

Drawings

In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings that are needed in the description of the embodiments or the prior art will be briefly described, and it is obvious that the drawings in the description below are some embodiments of the present invention, and other drawings can be obtained according to the drawings without inventive effort for a person skilled in the art.

Fig. 1 is a flowchart of a method for planning a driving track of an intersection according to an embodiment of the present invention;

FIG. 2 is a schematic view of an intersection scene provided by an embodiment of the present invention;

FIG. 3 is a schematic diagram of an optimal driving path according to an embodiment of the present invention;

FIG. 4 is a schematic diagram of another optimal driving path according to an embodiment of the present invention;

fig. 5 is a schematic structural diagram of a planning apparatus for a driving track at an intersection according to an embodiment of the present invention;

fig. 6 is a schematic structural diagram of an electronic device according to an embodiment of the present invention.

Detailed Description

The technical solutions of the present invention will be clearly and completely described in connection with the embodiments, and it is apparent that the described embodiments are some embodiments of the present invention, but not all embodiments. All other embodiments, which can be made by those skilled in the art based on the embodiments of the invention without making any inventive effort, are intended to be within the scope of the invention.

In the related art, a track planning method of an automatic driving vehicle at an intersection mainly comprises the steps of manually fitting a virtual lane line at the intersection, constructing a road network, and generating a virtual road center line as a virtual reference line for planning; such artificial construction of a virtual road network at intersections without lanes is not itself practical, resulting in a final map that is not "what you see is what you get". This is not obvious in the context of internal testing, but once it is desired to generalize, this approach is completely impractical because it is too much specific customization information, and when the mapping work is done by the grapher, it is not possible to require the grapher to provide all the customization information, thus causing the downstream modules that rely on these non-standard information to not function properly. In addition, even if a virtual lane is artificially constructed, the work is extremely complicated and complicated. Because in order to cover as many driving scenarios as possible, many possible virtual lanes need to be manually drawn at the intersection to connect different intersection entry and exit points. When the number of the road lanes is large and the traffic condition is complex, constructing the road network consumes a great deal of time and energy, and is quite easy to make mistakes, so that the efficiency of road junction track planning is low and the accuracy is poor.

Based on the above, the embodiment of the invention provides a method, a device and electronic equipment for planning a driving track of an intersection, and the technology can be applied to a scene of planning the driving track of an automatic driving vehicle at the intersection.

In order to facilitate understanding of the embodiment, the method for planning the intersection driving track disclosed by the embodiment of the invention is described in detail first; as shown in fig. 1, the method comprises the steps of:

step S102, obtaining coordinates of a current position of a vehicle at an intersection and coordinates of at least one target point of the vehicle after the vehicle passes through the intersection; wherein each target point has its corresponding travel priority; different target points correspond to different lanes.

The intersections can be intersections, T-junctions and the like; the current position may be understood as a current positioning position of the vehicle before passing through the intersection, where the current position is generally a position closer to the intersection, for example, if the vehicle needs to turn, the current position may be a position corresponding to when the vehicle needs to start to prepare for turning, etc.; the target point can be understood as a designated point on a lane after the vehicle passes through the intersection, and a rule for selecting the target point can be preset, for example, a point which is 10 meters away from the intersection on the center line of the lane after the vehicle passes through the intersection can be used as the target point, and the like; the number of the target points can be one or more, for example, if the vehicle passes through the intersection and has only one drivable lane, the number of the target points is one, and if the vehicle passes through the intersection and has a plurality of drivable lanes, one target point can be selected from each drivable lane, the number of the target points is a plurality of target points; the driving priority may be used to indicate the priority of the lane where the vehicle is driving on the target point after passing through the intersection. In practical implementation, when a driving track of a vehicle at an intersection needs to be planned, coordinates of a current position of the vehicle before passing through the intersection and coordinates of one or more target points corresponding to the vehicle after passing through the intersection are generally required to be acquired first.

And step S104, taking the current position as an initial node, and determining the coordinates of the expansion nodes adjacent to the initial node according to the kinematic characteristics of the vehicle by adopting a mixed A algorithm based on the coordinates of the current position.

The above Hybrid a algorithm may also be referred to as Hybrid a algorithm, which is a highly efficient path planning algorithm; in actual implementation, the coordinates of the expansion node adjacent to the initial node may be determined according to the vehicle kinematic expansion node from the current position of the vehicle at the intersection using a hybrid a-algorithm, for example, the coordinates of the current position, the expansion step length, the front-rear wheelbase of the vehicle, and the angle value of front wheel rotation of the vehicle at each expansion step may be determined according to the vehicle kinematic expansion node using a hybrid a-algorithm.

Step S106, judging whether a specified curve exists between the extended node and the current target point according to the order of the running priority of at least one target point from high to low; wherein the designated curve is a Reeds-Shepp curve or a Dubin curve.

The Reeds-Shepp curve and the Dubins curve can be used for indicating the shortest path of the vehicle, and the main difference between the Reeds-Shepp curve and the Dubins curve is whether the vehicle is allowed to reverse or not, and usually, whether the Reeds-Shepp curve exists or not can be confirmed in a reverse allowing scene, and whether the Dubins curve exists or not can be confirmed in a reverse not allowing scene. In actual implementation, since there may be a plurality of target points, it is possible to confirm whether the Reeds-Shepp curve or Dubin curve exists according to the order of the traveling priority of the target points from high to low; for example, if there are two target points, i.e., the target point 1 and the target point 2, the traveling priorities are high and low, respectively, if it is confirmed that there is a Reeds-Shepp curve or Dubins curve between the extended node and the target point 1 of high priority, the subsequent steps may be continued without confirming whether there is a Reeds-Shepp curve or Dubins curve between the extended node and the target point 2 of low priority, or it may be understood that when it is judged that there is a designated curve between the extended node and the current target point in the order of the traveling priorities of at least one target point from high to low, the subsequent steps may be continued as long as the designated curve appears.

Step S108, if a specified curve exists, determining an optimal running path of the vehicle based on the current target point, the extension node and the specified curve.

In actual implementation, if the designated curve appears, it can be considered that based on the extended node, the current target point corresponding to the designated curve can be searched, and starting from the current target point, the vehicle can travel back forward according to the designated curve and the extended node, so that the optimal travel path of the vehicle can be determined. The round of path searching is finished, and the searching efficiency of the optimal running path of the vehicle can be effectively improved through the mode.

Step S110, if no designated curve exists between the extended node and each target point, judging whether the distance between the extended node and the current target point is smaller than a preset threshold value according to the order of the running priority of at least one target point from high to low; and if the distance between the extended node and the current target point is smaller than the preset threshold value, starting from the current target point, sequentially backtracking forward based on the extended node, and determining the optimal running path of the vehicle.

The preset threshold value can be set according to actual requirements, and is not limited herein; if it is determined that there is no designated curve between the extended node and each target point, it may be continuously determined whether the distance between the extended node and the current target point is smaller than the preset threshold according to the order of the traveling priority of at least one target point from high to low, for example, if the traveling priorities of the extended node and the target point 1 with high priority are still being respectively high and low, if the distance between the extended node and the target point 1 with high priority is smaller than the preset threshold, then the subsequent steps may be continuously performed, it may not be required to confirm whether the distance between the extended node and the target point 2 with low priority is smaller than the preset threshold again, or it may be understood that, if the distance between the extended node and the current target point is smaller than the preset threshold according to the order of the traveling priority of at least one target point, it is determined whether the distance between the extended node and the current target point is smaller than the preset threshold, as long as the distance between the extended node and the target point is smaller than the preset threshold, the current target point is considered that the extended node has been extended, the current Hybrid a search is ended, and the vehicle travels from the current point to the current point, and finally, and the vehicle travels from the current position to the most point.

And step S112, if the distance between the extended node and each target point is not smaller than the preset threshold value, the extended node is taken as a new initial node, and the step of determining the coordinates of the extended node adjacent to the initial node according to the kinematic characteristics of the vehicle by adopting a hybrid A algorithm is repeatedly executed until the optimal running path of the vehicle is determined.

In actual implementation, if the distance between the extended node and each target point is not less than the preset threshold, it indicates that the extended node is unsuitable, or has not been extended to the current target point, then it is necessary to continue to confirm the next extended node, and the above steps are repeatedly performed to confirm whether an optimal travel path exists.

And step S114, optimizing the optimal running path by adopting a time elastic band algorithm, and determining the intersection running track of the vehicle.

The above time elastic band algorithm may also be referred to as TEB (Time Elastic Band ) algorithm; in actual implementation, the determined optimal running path of the vehicle does not generally include time information, and the vehicle cannot necessarily run according to the optimal running path, for example, a Reeds-Shepp curve or a Dubins curve is generally in a form of a combination of a straight line and an arc, and may not be smooth at a connection point, and at this time, a time elastic band algorithm may be used to perform optimization smoothing processing on the optimal running path to obtain an intersection running track of the vehicle, where the intersection running track is a track where the vehicle can actually run.

The method for planning the running track of the intersection obtains the coordinates of the current position of the vehicle at the intersection and the coordinates of at least one target point of the vehicle after passing through the intersection; the method comprises the steps of taking a current position as an initial node, adopting a mixed A-type algorithm based on the coordinates of the current position, and determining the coordinates of an expansion node adjacent to the initial node according to the kinematic characteristics of a vehicle; and sequentially judging whether a specified curve exists between the extended node and the current target point according to the order of the running priority of at least one target point from high to low. If a specified curve exists, determining an optimal running path of the vehicle based on the current target point, the extension node and the specified curve; if no designated curve exists between the extended node and each target point, sequentially judging whether the distance between the extended node and the current target point is smaller than a preset threshold according to the order of the running priority of at least one target point from high to low, if the distance between the extended node and the current target point is smaller than the preset threshold, starting from the current target point, sequentially tracing back forward based on the extended node, and determining the optimal running path of the vehicle; if the distance between the extended node and each target point is not smaller than the preset threshold value, the extended node is taken as a new initial node, and the step of determining the coordinates of the extended node adjacent to the initial node according to the kinematic characteristics of the vehicle by adopting a mixed A algorithm is repeatedly executed until the optimal running path of the vehicle is determined; and (3) optimizing the optimal running path by adopting a time elastic band algorithm, and determining the intersection running track of the vehicle. According to the method, the optimal running path of the vehicle at the intersection can be automatically searched by adopting a mixed A-type algorithm, and the searched optimal running path is optimized by adopting a time elastic band algorithm, so that the intersection running track of the vehicle capable of actually running is obtained, and the planning efficiency and accuracy of the intersection running track can be improved because a virtual road network is not required to be constructed manually.

The embodiment of the invention also provides another method for planning the crossing driving track, which is realized on the basis of the method of the embodiment; the method comprises the following steps:

step one, acquiring coordinates of a current position of a vehicle at an intersection and an initial path of the vehicle; the initial path is a path between a running start point and a running end point of the vehicle which is planned in advance.

In actual implementation, the coordinates of the current position of the vehicle at the intersection, which is typically the coordinate values in the world coordinate system, and global path information, which typically includes an initial path of the vehicle from the driving start point to the driving end point, which is planned in advance, may be acquired by an upstream module. The current position can be used as a starting point for subsequent track planning.

And step two, determining a target lane after the vehicle passes through the intersection based on the initial path, and determining the running priority of the target lane as a first priority.

Step three, if other lanes which are in the same direction as the target lane exist, determining the driving priority of the other lanes in sequence; wherein, the farther the distance between other lanes and the target lane, the lower the driving priority of other lanes.

The first priority corresponds to the highest running priority; in specific implementation, the running direction of the vehicle at the current intersection can be determined according to the initial path in the global path information, and the target lane after passing through the intersection is determined as the lane with the highest running priority. If there are other lanes in the same direction as the target lane, the lanes adjacent to the left and right of the target lane are taken as the lanes of the secondary driving priority, the lanes farther away from the target lane are taken as the lanes of the secondary driving priority, and the like until the driving priorities of all the same-direction lanes are confirmed.

Selecting target points from the target lane and other lanes respectively, and determining coordinates of at least one target point; the driving priority corresponding to each target point is the driving priority of the lane where the target point is located.

Respectively selecting target points from the target lane and other lanes in the same direction as the target lane, and determining the coordinates of each target point to obtain a target point set with a driving priority; for example, a point 10 meters away from the intersection on the lane center line of each lane after the vehicle passes through the intersection may be taken as the target point of the lane, or the like. The driving priority corresponding to each target point is the driving priority of the lane where the target point is located.

For example, referring to a schematic view of an intersection scene shown in fig. 2, the intersection is an intersection, a vehicle needs to turn right from a starting point at the intersection to pass through the intersection, two lanes in the same direction are arranged after turning, and a target point is selected from each lane, namely a target point 1 and a target point 2 in the graph.

And fifthly, acquiring a preset expansion step length, front and rear wheelbases of the vehicle and an angle value of front wheel rotation of the vehicle during each expansion step.

And step six, determining the coordinates of the expansion nodes adjacent to the initial node according to the vehicle kinematic characteristics by adopting a hybrid A algorithm according to the coordinates of the current position, the expansion step length, the front and rear wheelbase and the angle value.

For ease of understanding, continuing with the description of fig. 2, starting from the start point in fig. 2, the extended node is determined according to vehicle kinematics using a hybrid a-algorithm, which may be specifically implemented by the following expression:

wherein, (x) _i ，y _i ，β _i ) Is the coordinate of the ith expansion node based on the world coordinate system in the Cartesian coordinate system; x is x _i And y _i Coordinate values of the ith expansion node in the x-axis direction and the y-axis direction, beta _i An included angle between the head direction of the vehicle at the ith expansion node and the x-axis direction; s is the step length of each step expansion (corresponding to the expansion step length); l is the front-rear wheelbase of the vehicle; the sizing is the magnitude of the angle value (in radians) by which the front wheel of the vehicle rotates during each step of extension.

S, L and scheduling in the above expression are usually fixed parameters set in advance; in addition, it should be noted that, when i=0, the corresponding expansion node is the current position of the vehicle at the intersection, and the corresponding coordinates are the coordinates of the current position of the vehicle at the intersection, i.e. the coordinates of each expansion node can be determined in sequence according to the coordinates of the current position.

And step seven, if the intersection belongs to a scene allowing reversing, sequentially judging whether a Reeds-Shepp curve exists between the extension node and the current target point according to the order of the driving priority of at least one target point from high to low.

And step eight, if the intersection does not belong to a scene allowing reversing, sequentially judging whether a Dubins curve exists between the extended node and the current target point according to the order of the driving priority of at least one target point from high to low.

In actual implementation, whether a scene to which the intersection belongs allows reversing is preset, and if the scene is the scene which allows reversing, according to the order of the driving priority of at least one target point from high to low, sequentially calculating whether a Reeds-Shepp curve exists between the current extended node and the current target point; if the scene is not allowed to be backed up, sequentially calculating whether a Dubins curve exists between the current extended node and the current target point.

Step nine, if a specified curve exists, determining an optimal running path of the vehicle based on the current target point, the extension node and the specified curve.

Step ten, if no designated curve exists between the extended node and each target point, judging whether the distance between the extended node and the current target point is smaller than a preset threshold value according to the order of the running priority of at least one target point from high to low; and if the distance between the extended node and the current target point is smaller than the preset threshold value, starting from the current target point, sequentially backtracking forward based on the extended node, and determining the optimal running path of the vehicle.

And step eleven, if the distance between the extended node and each target point is not smaller than a preset threshold value, taking the extended node as a new initial node, and repeatedly executing the step of determining the coordinates of the extended node adjacent to the initial node according to the kinematic characteristics of the vehicle by adopting a mixed A algorithm until the optimal running path of the vehicle is determined.

And step twelve, acquiring an incomplete cost function, a speed cost function, an acceleration cost function, an obstacle collision-free cost function and a time cost function of the vehicle.

And thirteenth, weighting the incomplete cost function, the speed cost function, the acceleration cost function, the obstacle collision-free cost function and the time cost function based on preset weight coefficients to obtain an objective function.

And fourteen, solving an optimal solution of the objective function to optimize the optimal running path and determine the intersection running track of the vehicle.

And on the basis of the path searched by the mixed A method, optimizing and smoothing the path by using a TEB algorithm. The specific principle is as follows: TEB is an optimization method based on graph optimization. The optimization variable is a continuous coordinate point (x _i ，y _i ，β _i ) And a time interval Δt between coordinate points _i The method comprises the steps that the value range of i is 0-n, all coordinate points are added into a vertex set of a TEB ultrasonic image at time intervals, wherein all coordinate points comprise coordinates of the current position of a vehicle at an intersection and coordinates of all expansion nodes; deltaT _i Representing a time interval between an i-th expansion node and a subsequent expansion node, if i=0, representing a time interval between a current position of the slave vehicle at the intersection and the first expansion node; the Δti may be calculated from coordinate points, for example, by calculating a distance between two coordinate points based on coordinate values of two adjacent coordinate points, and then calculating a time interval between the two coordinate points based on a preset maximum speed allowed by the vehicle.

The objective function may be determined by the following formula: objective function f=ω ₁ f _{non-holonomic} +ω ₂ f _v +ω ₃ f _a +ω ₄ f _obstacle +ω ₅ f _t The method comprises the steps of carrying out a first treatment on the surface of the Wherein f _{non-holonomic} Representing an incomplete cost function of the vehicle; f (f) _v Representing a speed cost function of the vehicle; f (f) _a Representing an acceleration cost function of the vehicle; f (f) _obstacle A collision-free cost function representing an obstacle of the vehicle; f (f) _t Representing a time cost function of the vehicle; ω1, ω2, ω3, ω4, and ω5 are weight coefficients corresponding to the respective cost functions, respectively. The optimal solution for the objective function is typically the solution when the objective function is at a minimum.

In particular, the non-integrity of the vehicleCost function f _{non-holonomic} Is determined by the following formula:

Specifically, referring to an optimal driving path schematic diagram shown in fig. 3, the schematic diagram includes an ith expansion node and an (i+1) th expansion node that are adjacent to each other, and the non-integrity constraint condition of the vehicle may be set as follows:

the above-described non-integrity constraint is understood to mean that if the vehicle is turning or moving over an arc of a circle, it is required that the vehicle can only move in the direction toward which the vehicle body is facing, and that the i-th expansion node and the i+1-th expansion node of the vehicle fall on the same arc of a circle, the turning radii of the vehicle are equal. Based on the non-integrity constraint, a non-integrity cost function f can be made _{non-holonomic} The value of (2) is as small as possible to obtain an incomplete cost function f _{non-holonomic} Is set to the optimum value of (2).

The speed cost function of the vehicle is determined by the following formula:

wherein,

The speed constraint of the vehicle at the ith expansion node may be set to v _min ≤v _i ≤v _max The method comprises the steps of carrying out a first treatment on the surface of the I.e. v which is desired to be calculated during the optimization process _i Satisfying this constraint to make the velocity cost function f _v The value of (2) is as small as possible to obtain a speed cost function f _v Is set to the optimum value of (2).

The acceleration cost function of the vehicle is determined by the following formula:

wherein,

representing acceleration of the vehicle at the i-th expansion node; v _i Representing the speed of the vehicle at the ith expansion node; v _i+1 Representing the speed of the vehicle at the i+1th expansion node; deltaT _i Representing a time interval between an i-th expansion node and an i+1th expansion node; deltaT _i+1 Representing the time interval between the i+1th expansion node and the i+2th expansion node.

The acceleration constraint of the vehicle at the ith expansion node may be set to a _min ≤a _i ≤a _max I.e. a which is desired to be calculated during the optimization process _i This constraint condition is satisfied so that the value of the acceleration cost function fa is as small as possible, and the optimal value of the acceleration cost function fa is obtained.

The obstacle collision-free cost function of the vehicle is determined by the following formula:

Above-mentionedAnd->Usually, the coordinate value of the obstacle under the world coordinate system can be obtained in advance, and the obstacle information within the preset range from the ith expansion node is obtained from all the obstacle information; in the optimization process, x _i And y _i Is dynamically changed, so f _obstacle Is also dynamically changing; referring to another optimal driving path diagram shown in fig. 4; the map comprises an ith-1 expansion node, an ith expansion node and an (i+1) th expansion node which are adjacent, wherein obstale_j represents a jth obstacle within a preset range from the ith expansion node, d_min is a preset minimum distance between a hoped vehicle and the obstacle, for example, d_min is 2 meters and the like. / >

The time cost function of the vehicle is determined by the following formula:

wherein DeltaT _i Representing the time interval between the i-th expansion node and the i+1th expansion node. During the optimization process, deltaT _i Is dynamically changed, it is desired to calculate f _t As small as possible to obtain a time cost function f _t Is set to the optimum value of (2).

In the above formulas, when i=0, the corresponding expansion node is the current position of the vehicle at the intersection, and the corresponding coordinate is the coordinate of the current position of the vehicle at the intersection.

In actual implementation, a G2O (General Graph Optimization, a C++ framework for optimizing a nonlinear error function) framework can be utilized to numerically solve a TEB optimization problem, namely, the optimization problem of a large-scale sparse matrix of the TEB constructed through the process is solved, an optimal solution, namely, a minimum value, of an objective function is calculated so as to achieve real-time requirements, and finally, a smooth and collision-free track from a starting point to multiple target points in a Cartesian coordinate system is obtained.

The embodiment discloses a complete frame scheme for the trajectory planning of an automatic driving vehicle in an intersection scene independent of a virtual road network, wherein a hybrid A-type searching method is adopted to automatically search a driving path of the automatic driving vehicle in the intersection, and in the hybrid A-type searching method, searching from a starting point to multiple target points is supported; the virtual road network is not required to be constructed artificially, and only a starting point and an end point are required to be provided, so that a path which can be driven by an automatic driving vehicle can be obtained in real time, and repeated and tedious work related to map drawing is reduced; the scheme supports a mode that a plurality of target lanes are taken as the end points, so that the optimal driving route from the current position to the crossing can be obtained, and the method is not limited to a single lane which can only reach the crossing. The route selection is more intelligent, so that an automatic driving vehicle can select the optimal way of passing through the intersection in real time according to actual conditions, and the method is closer to the driving logic of a person; in addition, the method adopts the TEB method to optimize on the basis of the searched path, and a smoother track is further obtained. The G2O framework is utilized to solve the optimization problem, so that the solving speed is high, the real-time requirement of the automatic driving vehicle can be better met, the time consumption of solving the problem is reduced, and the real-time requirement of the track planning of the automatic driving vehicle can be better met. In addition, unstructured road scenes outside the intersections, such as high-speed driving from a main road into an emergency lane scene, and the like, have good expansibility.

By adopting the scheme, in simulation test, the automatic driving vehicle can ensure better trafficability at the intersection, can avoid the obstacle better for the scene with the static obstacle, and can process the game with the dynamic obstacle to a certain extent.

The embodiment of the invention provides a planning device for a driving track of an intersection, as shown in fig. 5, the device comprises: an acquisition module 50 for acquiring coordinates of a current position of the vehicle at the intersection and coordinates of at least one target point after the vehicle passes through the intersection; wherein each target point has its corresponding travel priority; different target points correspond to different lanes; the first determining module 51 is configured to determine, based on the coordinates of the current position, coordinates of an expansion node adjacent to the initial node according to the kinematic characteristics of the vehicle by using a hybrid a-x algorithm with the current position as the initial node; a judging module 52, configured to sequentially judge whether a specified curve exists between the extended node and the current target point in an order of from high to low of the traveling priority of at least one target point; wherein the designated curve is a Reeds-Shepp curve or a Dubin curve; a second determining module 53 for determining an optimal travel path of the vehicle based on the current target point, the extension node, and the specified curve if the specified curve exists; a third determining module 54, configured to sequentially determine whether the distance between the extended node and the current target point is smaller than a preset threshold according to the order of the traveling priority of at least one target point from high to low if no specified curve exists between the extended node and each target point, and sequentially trace back from the current target point to determine an optimal traveling path of the vehicle based on the extended node if the distance between the extended node and the current target point is smaller than the preset threshold; a repeated execution module 55, configured to repeatedly execute the step of determining coordinates of the extended node adjacent to the initial node according to the kinematic characteristics of the vehicle by using the hybrid a algorithm with the extended node as a new initial node if the distance between the extended node and each target point is not less than a preset threshold, until an optimal driving path of the vehicle is determined; and the optimization processing module 56 is used for optimizing the optimal running path by adopting a time elastic band algorithm and determining the intersection running track of the vehicle.

The planning device for the road junction running track acquires the coordinates of the current position of the vehicle at the road junction and the coordinates of at least one target point of the vehicle after passing through the road junction; the method comprises the steps of taking a current position as an initial node, adopting a mixed A-type algorithm based on the coordinates of the current position, and determining the coordinates of an expansion node adjacent to the initial node according to the kinematic characteristics of a vehicle; and sequentially judging whether a specified curve exists between the extended node and the current target point according to the order of the running priority of at least one target point from high to low. If a specified curve exists, determining an optimal running path of the vehicle based on the current target point, the extension node and the specified curve; if no designated curve exists between the extended node and each target point, sequentially judging whether the distance between the extended node and the current target point is smaller than a preset threshold according to the order of the running priority of at least one target point from high to low, if the distance between the extended node and the current target point is smaller than the preset threshold, starting from the current target point, sequentially tracing back forward based on the extended node, and determining the optimal running path of the vehicle; if the distance between the extended node and each target point is not smaller than the preset threshold value, the extended node is taken as a new initial node, and the step of determining the coordinates of the extended node adjacent to the initial node according to the kinematic characteristics of the vehicle by adopting a mixed A algorithm is repeatedly executed until the optimal running path of the vehicle is determined; and (3) optimizing the optimal running path by adopting a time elastic band algorithm, and determining the intersection running track of the vehicle. The device can automatically search the optimal running path of the vehicle at the intersection by adopting a mixed A-type algorithm, and optimize the searched optimal running path by adopting a time elastic band algorithm to obtain the intersection running path of the vehicle capable of actually running.

Further, the acquisition module is further configured to: acquiring coordinates of a current position of a vehicle at an intersection and an initial path of the vehicle; the initial path is a path between a running start point and a running end point of a vehicle which is planned in advance; determining a target lane after a vehicle passes through an intersection based on the initial path, and determining the driving priority of the target lane as a first priority; if other lanes in the same direction as the target lane exist, sequentially determining the driving priority of the other lanes; the farther the distance between the other lanes and the target lane is, the lower the driving priority of the other lanes is; selecting target points from the target lane and other lanes respectively, and determining coordinates of at least one target point; the driving priority corresponding to each target point is the driving priority of the lane where the target point is located.

Further, the first determining module is further configured to: acquiring a preset expansion step length, front and rear wheelbases of a vehicle and an angle value of front wheel rotation of the vehicle during each expansion step; and determining the coordinates of the expansion nodes adjacent to the initial node according to the vehicle kinematic characteristics by adopting a mixed A-based algorithm according to the coordinates of the current position, the expansion step length, the front and rear wheelbase and the angle value.

Further, the judging module is further configured to: if the intersection belongs to a scene allowing reversing, sequentially judging whether a Reeds-Shepp curve exists between the extension node and the current target point according to the order of the running priority of at least one target point from high to low; if the intersection does not belong to the scene allowing reversing, judging whether a Dubin curve exists between the extension node and the current target point according to the order of the running priority of at least one target point from high to low.

Further, the optimization processing module is further configured to: acquiring an incomplete cost function, a speed cost function, an acceleration cost function, an obstacle collision-free cost function and a time cost function of a vehicle; weighting the incomplete cost function, the speed cost function, the acceleration cost function, the obstacle collision-free cost function and the time cost function based on preset weight coefficients to obtain an objective function; and solving an optimal solution of the objective function to optimize the optimal running path and determine the intersection running track of the vehicle.

wherein,(x _i ，yi，β _i ) Coordinates of an ith expansion node; (x) _i+1 ，y _i+1 ，β _i+1 ) Is the coordinates of the i+1th expansion node.

wherein,

wherein x is _i And y _i Coordinate values of the ith expansion node in the x-axis direction and the y-axis direction respectively; x is x _obstaclej And y _obstaclej Coordinate values of the jth obstacle in the x-axis direction and the y-axis direction in a preset range of the ith expansion node respectively; d, d _min Is a preset interval threshold value between the vehicle and the obstacle.

The device for planning the intersection driving track provided by the embodiment of the invention has the same implementation principle and the same technical effects as the embodiment of the method for planning the intersection driving track, and for the sake of brief description, the corresponding content in the embodiment of the method for planning the intersection driving track can be referred to.

The embodiment of the present invention further provides an electronic device, as shown in fig. 6, where the electronic device includes a processor 130 and a memory 131, where the memory 131 stores machine executable instructions that can be executed by the processor 130, and the processor 130 executes the machine executable instructions to implement the above-mentioned intersection driving track planning method.

Further, the electronic device shown in fig. 6 further includes a bus 132 and a communication interface 133, and the processor 130, the communication interface 133, and the memory 131 are connected through the bus 132.

The memory 131 may include a high-speed random access memory (RAM, random Access Memory), and may further include a non-volatile memory (non-volatile memory), such as at least one magnetic disk memory. The communication connection between the system network element and at least one other network element is implemented via at least one communication interface 133 (which may be wired or wireless), and may use the internet, a wide area network, a local network, a metropolitan area network, etc. Bus 132 may be an ISA bus, a PCI bus, an EISA bus, or the like. The buses may be divided into address buses, data buses, control buses, etc. For ease of illustration, only one bi-directional arrow is shown in FIG. 6, but not only one bus or type of bus.

The processor 130 may be an integrated circuit chip with signal processing capabilities. In implementation, the steps of the above method may be performed by integrated logic circuitry in hardware or instructions in software in processor 130. The processor 130 may be a general-purpose processor, including a central processing unit (Central Processing Unit, CPU for short), a network processor (Network Processor, NP for short), etc.; but also digital signal processors (Digital Signal Processor, DSP for short), application specific integrated circuits (Application Specific Integrated Circuit, ASIC for short), field-programmable gate arrays (Field-Programmable Gate Array, FPGA for short) or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components. The disclosed methods, steps, and logic blocks in the embodiments of the present invention may be implemented or performed. A general purpose processor may be a microprocessor or the processor may be any conventional processor or the like. The steps of the method disclosed in connection with the embodiments of the present invention may be embodied directly in the execution of a hardware decoding processor, or in the execution of a combination of hardware and software modules in a decoding processor. The software modules may be located in a random access memory, flash memory, read only memory, programmable read only memory, or electrically erasable programmable memory, registers, etc. as well known in the art. The storage medium is located in the memory 131, and the processor 130 reads the information in the memory 131, and in combination with its hardware, performs the steps of the method of the foregoing embodiment.

The embodiment of the invention also provides a machine-readable storage medium, which stores machine-executable instructions that, when being called and executed by a processor, cause the processor to implement the intersection driving track planning method, and the specific implementation can be referred to the method embodiment and will not be repeated herein.

The method, the device and the computer program product of the electronic device for planning the intersection driving track provided by the embodiment of the invention comprise a computer readable storage medium storing program codes, and instructions included in the program codes can be used for executing the method in the previous method embodiment, and specific implementation can be referred to the method embodiment and will not be repeated here.

The functions, if implemented in the form of software functional units and sold or used as a stand-alone product, may be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention may be embodied essentially or in a part contributing to the prior art or in a part of the technical solution in the form of a software product stored in a storage medium, comprising several instructions for causing a computer device (which may be a personal computer, a server, a network device, etc.) to perform all or part of the steps of the method of the embodiments of the present invention. And the aforementioned storage medium includes: a U-disk, a removable hard disk, a Read-Only Memory (ROM), a random access Memory (RAM, random Access Memory), a magnetic disk, or an optical disk, or other various media capable of storing program codes.

Finally, it should be noted that: the above embodiments are only for illustrating the technical solution of the present invention, and not for limiting the same; although the invention has been described in detail with reference to the foregoing embodiments, it will be understood by those of ordinary skill in the art that: the technical scheme described in the foregoing embodiments can be modified or some or all of the technical features thereof can be replaced by equivalents; such modifications and substitutions do not depart from the spirit of the invention.

Claims

1. The method for planning the intersection driving track is characterized by comprising the following steps:

acquiring coordinates of a current position of a vehicle at an intersection and coordinates of at least one target point of the vehicle after the vehicle passes through the intersection; wherein each target point has its corresponding driving priority; different target points correspond to different lanes;

the current position is taken as an initial node, a mixed A algorithm is adopted based on the coordinates of the current position, and the coordinates of expansion nodes adjacent to the initial node are determined according to the kinematic characteristics of the vehicle;

sequentially judging whether a specified curve exists between the extended node and the current target point according to the order of the running priority of the at least one target point from high to low; wherein the specified curve is a Reeds-Shepp curve or a Dubin curve;

Determining an optimal travel path of the vehicle based on the current target point, the extension node, and the specified curve if the specified curve exists;

if the designated curve does not exist between the extension node and each target point, sequentially judging whether the distance between the extension node and the current target point is smaller than a preset threshold value according to the order of the running priority of the at least one target point from high to low; if the distance between the extended node and the current target point is smaller than the preset threshold value, starting from the current target point, determining an optimal running path of the vehicle based on sequential forward backtracking of the extended node;

if the distance between the extended node and each target point is not smaller than a preset threshold value, taking the extended node as a new initial node, and repeatedly executing the steps of adopting a mixed A algorithm to determine the coordinates of the extended node adjacent to the initial node according to the kinematic characteristics of the vehicle until the optimal running path of the vehicle is determined;

adopting a time elastic band algorithm to optimize the optimal running path and determining the intersection running track of the vehicle;

The step of optimizing the optimal running path by adopting a time elastic band algorithm and determining the crossing running track of the vehicle comprises the following steps:

acquiring an incomplete cost function, a speed cost function, an acceleration cost function, an obstacle collision-free cost function and a time cost function of the vehicle;

weighting the incomplete cost function, the speed cost function, the acceleration cost function, the obstacle collision-free cost function and the time cost function based on preset weight coefficients to obtain an objective function;

solving an optimal solution of the objective function to optimize the optimal running path and determine the intersection running track of the vehicle;

the objective function is determined by the following formula:

objective function f=ω ₁ f _{non-holonomic} +ω ₂ f _v +ω ₃ f _a +ω ₄ f _obstacle +ω ₅ f _t ；

Wherein f _{non-holonomic} Representing an incomplete cost function of the vehicle; f (f) _v Representing a speed cost function of the vehicle; f (f) _a Representing an acceleration cost function of the vehicle; f (f) _obstacle A collision-free cost function representing an obstacle of the vehicle; f (f) _t Representing a time cost function of the vehicle; ω1, ω2, ω3, ω4 and ω5 are weight coefficients corresponding to the cost functions respectively;

non-integrity cost function f of the vehicle _{non-holonomic} Is determined by the following formula:

wherein,(x _i ，y _i ，β _i ) Coordinates of an ith expansion node; (x) _i+1 ，y _i+1 ，β _i+1 ) Coordinates of the (i+1) th expansion node;

the speed cost function of the vehicle is determined by the following formula:

wherein,

representing a speed of the vehicle at an i-th expansion node; (x) _i ，y _i ，β _i ) Coordinates of an ith expansion node; (x) _i+1 ，y _i+1 ，β _i+1 ) Coordinates of the (i+1) th expansion node; v _min A preset first speed threshold value; v _max A preset second speed threshold value; and v _min ＜v _max ；

wherein,

representing acceleration of the vehicle at an ith expansion node; v _i Representing a speed of the vehicle at an i-th expansion node; v _i+1 Representing a speed of the vehicle at an i+1th expansion node; deltaT _i Representing a time interval between an i-th expansion node and an i+1th expansion node; deltaT _i+1 Representing a time interval between the i+1th expansion node and the i+2th expansion node;

wherein x is _i And y _i Coordinate values of the ith expansion node in the x-axis direction and the y-axis direction respectively;and->Coordinate values of the jth obstacle in the x-axis direction and the y-axis direction in a preset range of the ith expansion node respectively; d, d _min A preset interval threshold value between the vehicle and the obstacle;

the time cost function of the vehicle is determined by the following formula:

wherein DeltaT _i Representing the time interval between the i-th expansion node and the i+1th expansion node.

2. The method of claim 1, wherein the step of obtaining coordinates of a current location of the vehicle at the intersection and coordinates of at least one target point of the vehicle after passing through the intersection comprises:

acquiring coordinates of a current position of the vehicle at an intersection and an initial path of the vehicle; the initial path is a path between a running start point and a running end point of the vehicle which is planned in advance;

determining a target lane after the vehicle passes through the intersection based on the initial path, and determining the driving priority of the target lane as a first priority;

if other lanes which are in the same direction as the target lane exist, sequentially determining the driving priority of the other lanes; the farther the distance between the other lanes and the target lane is, the lower the driving priority of the other lanes is;

selecting the target point from the target lane and the other lanes respectively, and determining the coordinate of the at least one target point; the driving priority corresponding to each target point is the driving priority of the lane where the target point is located.

3. The method of claim 1, wherein the step of determining coordinates of an extended node adjacent to the initial node in terms of vehicle kinematic characteristics using a hybrid a-algorithm based on the coordinates of the current location with the current location as the initial node comprises:

acquiring a preset expansion step length, front and rear wheelbases of the vehicle and an angle value of front wheel rotation of the vehicle during each expansion step;

and determining the coordinates of the expansion nodes adjacent to the initial node according to the vehicle kinematic characteristics by adopting a mixed A-type algorithm according to the coordinates of the current position, the expansion step length, the front and rear wheelbase and the angle value.

4. The method according to claim 1, wherein the step of sequentially judging whether a specified curve exists between the extended node and the current target point in order of the traveling priority of the at least one target point from high to low comprises:

if the intersection belongs to a scene allowing reversing, sequentially judging whether the Reeds-Shepp curve exists between the extended node and the current target point according to the order of the running priority of the at least one target point from high to low;

And if the intersection does not belong to the scene allowing reversing, sequentially judging whether the Dubin curve exists between the extended node and the current target point according to the order of the running priority of the at least one target point from high to low.

5. The method of claim 1, wherein the corresponding extension node when i = 0 is the current location of the vehicle at the intersection.

6. A device for planning a travel path at an intersection, the device comprising:

the system comprises an acquisition module, a control module and a control module, wherein the acquisition module is used for acquiring the coordinates of the current position of a vehicle at an intersection and the coordinates of at least one target point of the vehicle after the vehicle passes through the intersection; wherein each target point has its corresponding driving priority; different target points correspond to different lanes;

the first determining module is used for determining the coordinates of an expansion node adjacent to the initial node according to the vehicle kinematic characteristics by adopting a mixed A algorithm based on the coordinates of the current position by taking the current position as the initial node;

the judging module is used for sequentially judging whether a specified curve exists between the extended node and the current target point according to the order of the running priority of the at least one target point from high to low; wherein the specified curve is a Reeds-Shepp curve or a Dubin curve;

A second determining module configured to determine an optimal travel path of the vehicle based on the current target point, the extension node, and the specified curve, if the specified curve exists;

a third determining module, configured to sequentially determine whether a distance between the extended node and the current target point is smaller than a preset threshold according to an order of a driving priority of the at least one target point from high to low if the specified curve does not exist between the extended node and each target point; if the distance between the extended node and the current target point is smaller than the preset threshold value, starting from the current target point, determining an optimal running path of the vehicle based on sequential forward backtracking of the extended node;

a repeated execution module, configured to repeatedly execute the step of determining coordinates of an extended node adjacent to the initial node according to the kinematic characteristics of the vehicle by using a hybrid a algorithm with the extended node as a new initial node if the distance between the extended node and each target point is not less than a preset threshold value until an optimal running path of the vehicle is determined;

the optimization processing module is used for optimizing the optimal running path by adopting a time elastic band algorithm and determining the intersection running track of the vehicle;

The optimization processing module is also used for: acquiring an incomplete cost function, a speed cost function, an acceleration cost function, an obstacle collision-free cost function and a time cost function of a vehicle; weighting the incomplete cost function, the speed cost function, the acceleration cost function, the obstacle collision-free cost function and the time cost function based on preset weight coefficients to obtain an objective function; solving an optimal solution of the objective function to optimize the optimal running path and determine the intersection running track of the vehicle;

the objective function is determined by the following formula: objective function f=ω ₁ f _{non-holonomic} +ω ₂ f _v +ω ₃ f _a +ω ₄ f _obstacle +ω ₅ f _t The method comprises the steps of carrying out a first treatment on the surface of the Wherein f _{non-holonomic} Representing an incomplete cost function of the vehicle; f (f) _v Representing a speed cost function of the vehicle; f (f) _a Representing an acceleration cost function of the vehicle; f (f) _obstacle A collision-free cost function representing an obstacle of the vehicle; f (f) _t Representing a time cost function of the vehicle; ω1, ω2, ω3, ω4 and ω5 are weight coefficients corresponding to the cost functions respectively;

non-integrity cost function f of vehicle _{non-holonomic} Is determined by the following formula:

the speed cost function of the vehicle is determined by the following formula:

Wherein,

representing the speed of the vehicle at the ith expansion node; (x) _i ，y _i ，β _i ) Coordinates of an ith expansion node; (x) _i+1 ，y _i+1 ，β _i+1 ) Coordinates of the (i+1) th expansion node; v _min A preset first speed threshold value; v _max A preset second speed threshold value; and v _min ＜v _max ；

wherein,

representing acceleration of the vehicle at the i-th expansion node; v _i Representing the speed of the vehicle at the ith expansion node; v _i+1 Representing the speed of the vehicle at the i+1th expansion node; deltaT _i Representing a time interval between an i-th expansion node and an i+1th expansion node; deltaT _i+1 Representing a time interval between the i+1th expansion node and the i+2th expansion node;

the obstacle-free collision cost function of a vehicle is determined by the following formula:

the time cost function of the vehicle is determined by the following formula:where Δti represents the time interval between the i-th expansion node and the i+1th expansion node.

7. An electronic device comprising a processor and a memory, the memory storing machine executable instructions executable by the processor, the processor executing the machine executable instructions to implement the method of planning an intersection travel trajectory of any one of claims 1-5.

8. A machine-readable storage medium storing machine-executable instructions which, when invoked and executed by a processor, cause the processor to implement the method of planning an intersection travel trajectory of any one of claims 1-5.