CN118209127B - Path optimization method, path optimization device, electronic equipment and storage medium - Google Patents

Abstract

The present application relates to the field of path planning technology, and provides a path optimization method, device, electronic device and storage medium. The method obtains the navigation route map and obstacle information of the vehicle's current travel, rasterizes the navigation route map based on the static obstacle information in the obstacle information to obtain a first-time predicted value map, updates the first-time predicted value map based on the dynamic obstacle information in the obstacle information to obtain a second-time predicted value map, uses a local path planning algorithm to obtain a set of predicted trajectories of the vehicle in a drivable area, and uses the second-time predicted value map to determine the target trajectory from the predicted trajectory set to obtain an optimized path, enriches the constraints during local path planning, improves the efficiency of path optimization, and reduces the possibility of swinging back and forth after the optimization, improves the comfort of vehicle driving, and enhances the user experience.

Description

Path optimization method, device, electronic device and storage medium

Technical Field

The present application relates to the field of path planning technology, and in particular to a path optimization method, device, electronic device and storage medium.

Background Art

During the vehicle's automatic driving process, efficient path planning and path tracking optimization control help the vehicle achieve fast navigation tasks.

Currently, the commonly used path planning algorithm is the A* (A-Star) algorithm.

However, the A* algorithm is a direct search method for solving the shortest path in a static road network. In complex known local spaces, the traditional A* algorithm performs path planning and has poor dynamic obstacle avoidance effects. Furthermore, when improving the search controls and cost functions of the A* algorithm in related technologies, the topological constraints of lanes, traffic rules, and abnormal situations in the road network are not considered, resulting in the planned trajectory swinging back and forth, affecting the user experience.

Summary of the invention

In view of this, the embodiments of the present application provide a path optimization method, device, electronic device and storage medium to solve the problem in the prior art that the path optimization constraints are not perfect, resulting in complex calculations and the risk of collision in the optimized path.

In a first aspect of an embodiment of the present application, a path optimization method is provided, the method being used to optimize a vehicle navigation path, the method comprising:

Obtaining a navigation route map and obstacle information of the vehicle's current travel, where the obstacle is an obstacle in the navigation route, and the obstacle information includes at least the location information and attribute information of static obstacles, and the predicted trajectory of dynamic obstacles;

The navigation route map is rasterized, and each grid is assigned a value based on the location information and attribute information of the static obstacles to obtain a first-time prediction value map;

Based on the predicted trajectory of the dynamic obstacle, the first-time predicted value map is updated to obtain the second-time predicted value map, wherein the first-time predicted value map and the second-time predicted value map are used to indicate the distance between the vehicle and the obstacle at different times;

Determine a drivable area of the vehicle from time t to time t+1 based on the second time prediction value map;

Obtain the state information of the vehicle at time t and the target search step length, perform local path planning based on the state information of the vehicle at time t and the target search step length according to the dynamic constraints of the Ackerman chassis model, and obtain a set of predicted trajectories of the vehicle from time t to time t+1, where t is greater than 0 and less than the total driving time of the vehicle;

A target trajectory is determined from the set of predicted trajectories based on the second time prediction value map, and the target trajectory is used as an optimized path for the vehicle from time t to time t+1.

A second aspect of an embodiment of the present application provides a path optimization device, including:

An acquisition module is configured to acquire a navigation route map and obstacle information of the vehicle's current travel, wherein the obstacle is an obstacle in the navigation route, and the obstacle information includes at least position information and attribute information of static obstacles and predicted trajectory of dynamic obstacles;

A first map generation module is configured to perform rasterization processing on the navigation route map and assign values to each grid based on the location information and attribute information of the static obstacles to obtain a first-time prediction value map;

The second map generation module is configured to update the first-time predicted value map based on the predicted trajectory of the dynamic obstacle to obtain a second-time predicted value map, wherein the first-time predicted value map and the second-time predicted value map are used to indicate the distance between the vehicle and the obstacle at different times;

A determination module configured to determine a drivable area of the vehicle from time t to time t+1 based on the second time prediction value map;

A planning module is configured to obtain state information of the vehicle at time t and a target search step length, perform local path planning based on the state information of the vehicle at time t and the target search step length according to the dynamic constraints of the Ackerman chassis model, and obtain a set of predicted trajectories of the vehicle from time t to time t+1, where t is greater than 0 and less than the total driving time of the vehicle;

The optimization module is configured to determine a target trajectory from the set of predicted trajectories based on the second time prediction value map, and use the target trajectory as an optimized path of the vehicle from time t to time t+1.

According to a third aspect of an embodiment of the present application, an electronic device is provided, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor implements the steps of the above method when executing the computer program.

According to a fourth aspect of an embodiment of the present application, a computer-readable storage medium is provided, which stores a computer program, and when the computer program is executed by a processor, the steps of the above method are implemented.

Compared with the prior art, the embodiments of the present application have the following beneficial effects: the embodiments of the present application obtain the navigation route map and obstacle information of the vehicle's current travel, rasterize the navigation route map based on the static obstacle information in the obstacle information to obtain a first-time predicted value map, and update the first-time predicted value map based on the dynamic obstacle information in the obstacle information to obtain a second-time predicted value map, use a local path planning algorithm to obtain a set of predicted trajectories of the vehicle in the drivable area from time t to time t+1 that satisfy the dynamic constraints of the Ackerman chassis model, and use the second-time predicted value map to determine the target trajectory from the predicted trajectory set, and use the target trajectory as the optimized path of the vehicle from time t to time t+1, thereby enriching the constraints during local path planning, and obtaining a smooth trajectory through one calculation, thereby improving the efficiency of path optimization, and the optimized path reduces the possibility of swinging back and forth, thereby improving the driving comfort of the vehicle and the user experience.

BRIEF DESCRIPTION OF THE DRAWINGS

In order to more clearly illustrate the technical solutions in the embodiments of the present application, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

FIG1 is a flow chart of a path optimization method provided in an embodiment of the present application.

FIG. 2 is a schematic diagram of a navigation route map provided in an embodiment of the present application.

FIG3 is a schematic diagram of a map after dividing the regions according to the values of each grid in the second time predicted value map provided in an embodiment of the present application.

4 is a flow chart of a method for updating a first-time predicted value map based on a predicted trajectory of a dynamic obstacle to obtain a second-time predicted value map provided in an embodiment of the present application.

FIG5 is a schematic diagram of the Ackerman chassis model provided in an embodiment of the present application.

FIG6 is a flow chart of a method for determining a target trajectory from a set of predicted trajectories based on a second time predicted value map provided in an embodiment of the present application.

FIG. 7 is a flow chart of another path optimization method provided in an embodiment of the present application.

FIG8 is a schematic diagram of a path optimization device provided in an embodiment of the present application.

FIG. 9 is a schematic diagram of an electronic device provided in an embodiment of the present application.

DETAILED DESCRIPTION

In the following description, specific details such as specific system structures, technologies, etc. are provided for the purpose of illustration rather than limitation, so as to provide a thorough understanding of the embodiments of the present application. However, it should be clear to those skilled in the art that the present application may also be implemented in other embodiments without these specific details. In other cases, detailed descriptions of well-known systems, devices, circuits, and methods are omitted to prevent unnecessary details from obstructing the description of the present application.

A path optimization method and device according to an embodiment of the present application will be described in detail below with reference to the accompanying drawings.

As mentioned above, the A* algorithm, a commonly used path planning algorithm, is a direct search method for solving the shortest path in a static road network. In complex known local spaces, the traditional A* algorithm performs path planning, and the dynamic obstacle avoidance effect is poor. Furthermore, when improving the search control and cost function of the A* algorithm in the related art, the lane topology constraints, traffic rules constraints, and abnormal situation constraints in the road network are not considered, causing the planned trajectory to swing back and forth, affecting the user experience, and there is a risk of collision when there are no obstacle constraints during trajectory optimization. Furthermore, the A* algorithm in the related art usually adds an additional smoothing process after path planning to obtain a smooth trajectory. This processing method increases the amount of calculation.

In view of this, an embodiment of the present application provides a path optimization method, which obtains a navigation route map and obstacle information of the vehicle's current travel, rasterizes the navigation route map based on the static obstacle information in the obstacle information to obtain a first-time predicted value map, and updates the first-time predicted value map based on the dynamic obstacle information in the obstacle information to obtain a second-time predicted value map, uses a local path planning algorithm to obtain a set of predicted trajectories of the vehicle in a drivable area from time t to time t+1 that satisfies the dynamic constraints of the Ackerman chassis model, and uses the second-time predicted value map to determine a target trajectory from the predicted trajectory set, and uses the target trajectory as the optimized path of the vehicle from time t to time t+1, enriching the constraints during local path planning, and obtaining a smooth trajectory through one calculation, thereby improving the efficiency of path optimization, and the optimized path reduces the possibility of swinging back and forth, thereby improving the driving comfort of the vehicle and the user experience.

FIG1 is a flow chart of a path optimization method provided in an embodiment of the present application. As shown in FIG1 , the method includes the following steps:

In step S101, the navigation route map and obstacle information of the vehicle's current travel are obtained.

The obstacle is an obstacle in the navigation route, and the obstacle information includes at least the location information and attribute information of the static obstacle and the predicted trajectory of the dynamic obstacle.

In step S102, the navigation route map is rasterized, and each grid is assigned a value based on the location information and attribute information of the static obstacles to obtain a first-time predicted value map.

In step S103, the first time predicted value map is updated based on the predicted trajectory of the dynamic obstacle to obtain a second time predicted value map.

Among them, the first-time predicted value map and the second-time predicted value map are used to indicate the distance between the vehicle and the obstacle at different times.

In step S104, a drivable area of the vehicle from time t to time t+1 is determined based on the second time prediction value map.

In step S105, the state information of the vehicle at time t and the target search step length are obtained, and local path planning is performed based on the state information of the vehicle at time t and the target search step length according to the dynamic constraints of the Ackerman chassis model to obtain a set of predicted trajectories of the vehicle from time t to time t+1.

Among them, t is greater than 0 and less than the total driving time of the vehicle.

In step S106, a target trajectory is determined from the set of predicted trajectories based on the second time prediction value map, and the target trajectory is used as the optimized path of the vehicle from time t to time t+1.

In the embodiment of the present application, the method can be executed by a server or by a terminal with certain computing capabilities. For the convenience of description, the following description is based on an example of the method being executed by a server.

In an embodiment of the present application, the server may first obtain a navigation route map of the vehicle's current travel. In one example, the navigation route map of the vehicle's current travel may be first generated based on a navigation application, and then the navigation route map may be obtained from the navigation application. Furthermore, other methods may be used to obtain the navigation route map, which is not limited here. It should be noted that the map may be a high-precision map.

In an embodiment of the present application, the server may also obtain obstacle information. The obstacle is an obstacle in the navigation route. The obstacle information includes at least the location information and attribute information of static obstacles, and the predicted trajectory of dynamic obstacles. In one example, static obstacles may be, for example, the boundary of the road, the sidewalk and guardrail on the side of the road, virtual and real lane lines, the lane driving direction specified by the traffic signs of each lane, stop lines, zebra crossings, waiting areas, and lane connectivity. In another example, dynamic obstacles may be, for example, traffic participants at a specific moment, including other vehicles, pedestrians, etc.

Fig. 2 is a schematic diagram of a navigation route map provided in an embodiment of the present application. As shown in Fig. 2, the navigation route map includes the above-mentioned various types of static obstacles.

In the embodiment of the present application, the obtained navigation route map can be rasterized, and each grid can be assigned a value based on the location information and attribute information of the static obstacle to obtain a first-time predicted value map. The specific assignment method is described in detail below and will not be repeated here.

In an embodiment of the present application, the first time prediction value map can also be updated based on the predicted trajectory of the dynamic obstacle to obtain a second time prediction value map. Among them, the second time prediction value map is a time prediction value map. This is because, at different times of the vehicle's current travel, there are usually different traffic participants, or the positions of the same traffic participants may be different in the maps at different times. Therefore, it is necessary to perform different update processing on the first time prediction value map according to the traffic participant situation at different times, that is, the predicted trajectory of the dynamic obstacle, to obtain a series of second time prediction value maps. Furthermore, the first time prediction value map and the second time prediction value map are used to indicate the distance of the vehicle from the obstacle at different times.

In the embodiment of the present application, after the second time prediction value map is constructed, the server can determine the drivable area of the vehicle from time t to time t+1 based on the second time prediction value map. Furthermore, the server can also obtain the state information of the vehicle at time t and the target search step length, and perform local path planning based on the state information of the vehicle at time t and the target search step length according to the dynamic constraints of the Ackerman chassis model to obtain a set of predicted trajectories of the vehicle from time t to time t+1, where t is greater than 0 and less than the total driving time of the vehicle.

In an embodiment of the present application, after obtaining a set of predicted trajectories of the vehicle from time t to time t+1, the second time prediction value map can be further used to determine a target trajectory from the set of predicted trajectories, and use the target trajectory as the optimized path of the vehicle from time t to time t+1.

According to the technical solution provided in the embodiment of the present application, by obtaining the navigation route map and obstacle information of the vehicle's current travel, the navigation route map is rasterized based on the static obstacle information in the obstacle information to obtain a first-time predicted value map, and the first-time predicted value map is updated based on the dynamic obstacle information in the obstacle information to obtain a second-time predicted value map, and a local path planning algorithm is used to obtain a set of predicted trajectories of the vehicle in the drivable area from time t to time t+1 that satisfy the dynamic constraints of the Ackerman chassis model, and the second-time predicted value map is used to determine the target trajectory from the predicted trajectory set, and the target trajectory is used as the optimized path of the vehicle from time t to time t+1, thereby enriching the constraints during local path planning, and a smooth trajectory can be obtained through one calculation, thereby improving the efficiency of path optimization, and the optimized path reduces the possibility of swinging back and forth, thereby improving the driving comfort of the vehicle and the user experience.

In the embodiment of the present application, the attribute information of the static obstacle includes at least: the type information of the static obstacle and the traffic condition information of the static obstacle. Further, in the first time prediction value map, the grid value is negatively correlated with the passability degree of the static obstacle included in the grid.

The passability of the static obstacles included in each grid is determined in the following manner: in response to determining that the static obstacle is a collision type obstacle, the passability of the static obstacle is determined to be a first value; in response to determining that the static obstacle is a temporary passable obstacle, the passability of the static obstacle is determined to be a second value; in response to determining that the static obstacle is a conditional passable obstacle, the passability of the static obstacle is determined to be a third value; in response to determining that the static obstacle is an obstacle that does not need to be occupied during normal driving, the passability of the static obstacle is determined to be a fourth value; in response to determining that the static obstacle is a passable obstacle, the passability of the static obstacle is determined to be a fifth value. In which, the first value is less than the second value, the second value is less than the third value, the third value is less than the fourth value, and the fourth value is less than the fifth value.

That is, the server may first rasterize the acquired navigation route map, and then assign a value to each grid according to traffic rules and the degree of road driving danger. For example, the navigation route map may be divided into five drivable areas according to traffic rules and the degree of road driving danger, wherein the first drivable area may be an area where driving is allowed by traffic rules and a drivable area with a higher degree of danger, such as an area including guardrails, trees, poles, and sidewalks with curbs above a preset threshold. That is, the static obstacles in the first drivable area are collision type obstacles, and the degree of allowed passage of static obstacles in this area is the lowest, so its grid value may be set to the highest.

Furthermore, the second drivable area may be an area that can be used as a passage in special circumstances but cannot be occupied for a long time, such as an opposite lane, an intersection area, an emergency lane, a solid line portion of a lane that does not lead to the destination, a single solid line, a sidewalk with a curb below a preset threshold, etc. That is, the static obstacles in the second drivable area are obstacles that allow temporary passage, and the degree of passage allowed for static obstacles in this area is the second lowest, so its grid value can be set to the second highest.

The third drivable area may be an area that is passable when certain conditions specified in traffic rules are met, such as a sidewalk, a waiting area, a stop line, etc. That is, the static obstacles in the third drivable area are conditionally passable obstacles, and the passability of static obstacles in this area is one level lower than that in the second drivable area, so its grid value can be set to be one level lower than that in the second drivable area.

The fourth drivable area may be an area that does not need to be occupied under normal circumstances, such as a lane dotted line. When the vehicle is driving normally, it should try to keep driving in a straight line, so there is no need to occupy the lane dotted line for merging. That is, the static obstacles in the fourth drivable area are obstacles that do not need to be occupied during normal driving. The allowable passability of static obstacles in this area is one level lower than that of the third drivable area, so its grid value can be set to be one level lower than that of the third drivable area.

The fifth drivable area may be a permitted drivable area. The fifth drivable area needs to be a drivable area for the vehicle according to the navigation route and the lane connectivity relationship. For example, if the vehicle needs to turn right at a fork in the road according to the navigation route, then the area that cannot turn right or change lanes is not included. That is, the static obstacles in the fifth drivable area are permitted obstacles. The permitted degree of static obstacles in this area is the highest, so its grid value can be set to the lowest.

In one implementation, the grid value in the first drivable area may be set to 4, the grid value in the second drivable area may be set to 3, the grid value in the third drivable area may be set to 2, the grid value in the fourth drivable area may be set to 1, and the grid value in the fifth drivable area may be set to a value greater than 0 and less than 1. It should be noted that the closer the grid value in the fifth drivable area is to 0, the closer the area where the grid is located is to the center line of the lane.

Fig. 3 is a schematic diagram of a map after dividing the regions according to the values of each grid in the second time predicted value map provided by an embodiment of the present application. As shown in Fig. 3, the map can be divided into five different regions according to the values of each grid in the second time predicted value map, corresponding to the first drivable region, the second drivable region, the third drivable region, the fourth drivable region and the fifth drivable region respectively.

In this way, when generating the time prediction value map that represents the distance between the vehicle and the obstacle, the traffic rules constraints and lane topology constraints are taken into account, making the path optimization process more in line with the actual situation, and thus making the optimized path smoother, which can bring a more comfortable experience to users.

In the embodiment of the present application, as described above, the second time prediction value map is a time series map. FIG4 is a flow chart of a method for updating the first time prediction value map based on the predicted trajectory of a dynamic obstacle to obtain the second time prediction value map provided by the embodiment of the present application. As shown in FIG4, the method includes the following steps:

In step S401, the target predicted position of the dynamic obstacle at time t is obtained.

In step S402, a possible collision area of the dynamic obstacle at time t is calculated based on the speed of the dynamic obstacle at time t, the road condition information at time t, and the weather information at time t.

In step S403, in the first time prediction value map, the grid value at the target prediction position is updated to the maximum value, and the grid value in the possible collision area is updated to the second maximum value, so as to obtain a second time prediction value map.

In the embodiment of the present application, after generating the first time prediction value map, the server can further obtain the target predicted position of the dynamic obstacle at time t. The predicted position of the dynamic obstacle can be obtained by obtaining the state of the obstacle through sensors such as radar or camera, and then calculating its predicted trajectory, or by obtaining the positioning position of the dynamic obstacle and calculating its predicted trajectory based on the dynamic change of its positioning position, or by other methods, which are not limited in the embodiment of the present application.

In the embodiment of the present application, the server can also calculate the possible collision area of the dynamic obstacle at time t based on the speed of the dynamic obstacle at time t, the road condition information at time t, and the weather information at time t. Finally, in the first time prediction value map, the grid value at the target prediction position is updated to the maximum value, and the grid value in the possible collision area is updated to the second maximum value, so as to obtain the second time prediction value map.

For example, the server can obtain the predicted trajectory of traffic participants through the trajectory prediction module, thereby obtaining the position, speed, length, width and height of all traffic participants at a certain moment in the future, and set the grid value corresponding to the area of each traffic participant to the highest value (for example, 5) on the basis of the first-time predicted value map, and calculate the reserved distance around the traffic participant according to the current speed, road conditions and weather, and obtain the possible collision area defined by the reserved distance, and set the grid value corresponding to the possible collision area to the second highest value (for example, 4), thereby obtaining the second-time predicted value map at that moment.

In the embodiment of the present application, the server needs to determine the drivable area of the vehicle, and then use a local path planning algorithm in the drivable area to plan the optimal path of the vehicle from time t to time t + 1. The local path planning algorithm may be, for example, an A* algorithm.

In the embodiment of the present application, before determining the drivable area of the vehicle, the upper limit value of the area can be set first, and the upper limit value of the area is used to represent the maximum value of each grid in the drivable area of the vehicle. In one example, the drivable area when the vehicle is in the following state and the straight state can be set as the fifth drivable area, and the upper limit value H of the area can be 1 at this time; the drivable area when the vehicle is in the overtaking state and the lane changing state can be set as the fifth drivable area and the fourth drivable area, and H can be 2 at this time; the drivable area when the vehicle is at the intersection and the intersection signal light is red can be set as the fifth drivable area and the fourth drivable area, and H can be 2 at this time; the drivable area when the vehicle is at the intersection and the intersection signal light shows waiting to turn can be set as the fifth drivable area, the fourth drivable area and the third drivable area, and H can be 3 at this time; the drivable area when the vehicle is in the emergency avoidance state can be set as the fifth drivable area, the fourth drivable area, the third drivable area and the second drivable area, and H can be 4 at this time.

In the embodiment of the present application, an open set and a closed set can also be set, wherein the open set is the nodes corresponding to the grids that may be passed during path planning, and the closed set is the nodes corresponding to the grids that will not be passed during path planning and the grids that have been added to the optimized path. At this time, the drivable area of the vehicle from time t to time t+1 can be determined in the following way: determine the target open set of the vehicle from time t to time t+1 based on the second time prediction value map of time t and time t+1; determine the area composed of grids whose grid values corresponding to the nodes in the target open set are less than the upper limit value of the area, which is the drivable area of the vehicle from time t to time t+1.

In the embodiment of the present application, when performing local path planning, the vehicle state information and target search step length at time t can be first obtained. The vehicle state information at time t is Where _xt is the horizontal coordinate value of the vehicle at time t, _yt is the vertical coordinate value of the vehicle at time t, and _vt is the speed of the vehicle at time t. is the heading angle of the vehicle at time t, and _θt is the tire angle of the vehicle at time t. Further, the target search step is Δt, Δt=ΔT/N, ΔT is the preset search step, N represents the number of samples from time t to time t+1, and N is a positive integer greater than 1. Furthermore, the dynamic constraints of the Ackerman chassis model include constraints on vehicle speed and vehicle tire angle.

In the embodiment of the present application, when local path planning is performed in a drivable area based on the state information of the vehicle at time t and the target search step, M sampling combinations (Δv _t,i , Δθ _t,i ) can be first determined based on the state information of the vehicle at time t, where i is a positive integer from 1 to M and M is a positive integer greater than 1. Then, for each sampling combination, a path planning algorithm is used to predict M predicted trajectory sets of the vehicle at time t+1 that meet the dynamic constraints of the Ackerman chassis model in the drivable area according to the target search step.

Among them, each predicted trajectory set includes N predicted position nodes, each predicted position node corresponds to a search step, each predicted position node is obtained by searching the grid number corresponding to the node, and the vehicle state information corresponding to each predicted position node is x _t+1 is the horizontal coordinate value of the vehicle at time t+1, y _t+1 is the vertical coordinate value of the vehicle at time t+1, and v _t+1 is the speed of the vehicle at time t+1. is the heading angle of the vehicle at time t+1, and θ _t+1 is the tire angle of the vehicle at time t+1.

It should be noted that each predicted position node is obtained by searching the grid serial number corresponding to the node, which means that when obtaining the predicted position nodes in each predicted trajectory, each predicted position node can be first gridded to obtain the serial number of the grid where the node is located, and then searched in the open set according to the serial number to obtain the predicted position nodes in each predicted trajectory that are located in the open set.

That is to say, since each node uses vehicle status information It means that when searching for it, if we search for each parameter separately, the calculation amount is too large and the efficiency is too low. In view of this, we can grid each node and define each grid The size of is defined as (0.1, 0.1, 0.1, 0.1, 0.05), and then the node value is divided by the defined size and the result is rounded to get the sequence number of the grid where the current node is located. Next, the sequence number is used to search in the open combination and closed set, which can greatly reduce the amount of calculation and improve the search efficiency.

Fig. 5 is a schematic diagram of the Ackerman chassis model provided by an embodiment of the present application. As shown in Fig. 5, the wheelbase of the left and right wheels of the vehicle is represented by 2d, the wheelbase of the front and rear wheels is represented by l, the vehicle tire angle θ is represented, and the radius of the vehicle turning is represented by R.

In the embodiment of the present application, the vehicle status information corresponding to each predicted location node It can be determined as follows:

Wherein, n is a positive integer greater than 0 and less than N.

FIG6 is a flow chart of a method for determining a target trajectory from a set of predicted trajectories based on a second time predicted value map provided by an embodiment of the present application. As shown in FIG6 , the method includes the following steps:

In step S601, an evaluation function is generated based on the second time prediction value map.

In step S602, the cost value of each predicted position node in the predicted trajectory set is calculated using an evaluation function.

In step S603, the sum of the cost values of the predicted position nodes in each predicted trajectory is determined as the cost value of each predicted trajectory.

In step S604, the predicted trajectory with the minimum cost is determined as the target trajectory.

In the embodiment of the present application, when determining the target trajectory from the predicted trajectory set based on the second time predicted value map, an evaluation function can be first generated based on the second time predicted value map. Then, the evaluation function is used to calculate the cost value of each predicted position node in the predicted trajectory set that is located in the open set. Next, the sum of the cost values of each predicted position node in the open set in each predicted trajectory is determined as the cost value of each predicted trajectory. Finally, the predicted trajectory with the smallest cost value is determined as the target trajectory.

In the embodiment of the present application, the evaluation function includes a first evaluation item and a second evaluation item. The first evaluation item is the sum of the cost of the vehicle from the position of the vehicle at the starting time to the position of the vehicle at time t and the cost of the vehicle from the position of the vehicle at time t to the predicted position node to be evaluated. The second evaluation item is the estimated cost of the vehicle from the predicted position node to be evaluated to the target node, and the target node is the terminal position in the predicted trajectory where the predicted position node to be evaluated is located.

That is to say, the evaluation function may be, for example, f(t+1), and f(t+1)=g(t+1)+h(t+1), where g(t+1) is the first evaluation item. g(t) represents the cost of the vehicle from its starting position to its position at time t. represents the cost value of the vehicle from the position of the vehicle at time t to the predicted position node to be evaluated; Among them, k ₁ , k ₂ and k ₃ are weight values, which can be set according to actual needs, d(x,y) is the distance cost, V(v) is the speed cost, is the heading cost. The closer d(x,y) is to 0, the larger k ₂ and k ₃ are.

Furthermore, the cost value of each predicted position node in the prediction trajectory set that is located in the open set is calculated using an evaluation function, which can be: using the evaluation function to calculate a first evaluation item value of the predicted position node to be evaluated; using the evaluation function to calculate a second evaluation item value of the predicted position node to be evaluated; taking the sum of the first evaluation item value and the second evaluation item value as the cost value of the predicted position node to be evaluated, and the predicted position node to be evaluated is any node among the predicted position nodes in the prediction trajectory set that are located in the open set.

In the embodiment of the present application, after using the evaluation function to calculate the cost value of each predicted location node in the open set in the predicted trajectory set, if the first evaluation item of the predicted location node to be evaluated is greater than or equal to the regional upper limit value, the predicted location node to be evaluated can be added to the closed set, and the predicted trajectory where the predicted location node to be evaluated is located is discarded, or the predicted trajectory where the predicted location node to be evaluated is located is regenerated. Conversely, if the first evaluation item of the predicted location node to be evaluated is less than the regional upper limit value, the predicted location node to be evaluated is added to the open set, and the time attribute of time t+1 is added to the predicted location node to be evaluated.

In the embodiment of the present application, each node in the open set contains pheromones. Further, after adding the predicted position node to be evaluated to the closed set, the target pheromone of the predicted position node to be evaluated added to the closed set can also be determined; then, the pheromone of each node in the open set whose distance to the predicted position node to be evaluated added to the closed set is less than a preset distance threshold is subtracted from the target pheromone, and the node corresponding to the position of the vehicle at time t is increased by the target pheromone.

In an embodiment of the present application, the second evaluation item value is obtained by at least weighted summing up the distance estimated cost, speed estimated cost and heading estimated cost of the vehicle from the predicted position node to be evaluated to the target node.

FIG7 is a flow chart of another path optimization method provided by an embodiment of the present application. As shown in FIG7, the high-precision map can be firstly rasterized and the region can be divided to obtain the first-time predicted value map; then the predicted trajectory of the traffic participant predicted by the trajectory prediction module is used to update the first-time predicted value map to obtain the second-time predicted value map, and then the drivable area and short-term target node are determined by the decision module, and the optimal path of the vehicle from time t to time t+1 is predicted by combining the local path planning algorithm with the second-time predicted value map.

That is to say, the decision-making module can be used to determine the drivable area and the short-term target node, and the trajectory planning module can be used to calculate the trajectory in the drivable area. In order to make the calculated trajectory more comfortable and reduce the amount of calculation, the preset search step size can be set to ΔT. First, the optimal search node is obtained in the open set. Assuming that the time from the start time to the current time of the optimal node is t, the step size is evenly divided into n parts, one part is Δt, then ΔT=n*Δt, and the search space is the M changes in speed and steering wheel angle within Δt (Δv, Δθ). For each sampling combination, the corresponding vehicle state at the i+1th moment is obtained, including at least the vehicle speed as v _t+1 and the vehicle tire angle as θ _t+1 . N trajectory points will be generated, and the vehicle position and vehicle heading at the t+1th moment are calculated as Where x and y are the positions, φ is the vehicle heading, and the Ackerman chassis vehicle dynamics constraints are considered.

Where n∈integer (0,N). M nodes are generated at time t and time t+1. The M nodes at time t+1 are traversed to determine whether they are in the closed set. In some embodiments, in order to facilitate searching in the closed set and the open set, Gridding, get the number of the grid where the current node is located, and then use the number to find the node. Next, calculate the cost of each node according to the evaluation function. The evaluation function is f(t+1)=g(t+1)+h(t+1). f(t+1) is the estimated cost of reaching the target node from the starting node through a node at time t+1. g(t+1) is the actual cost from the starting node to a node at time t+1, expressed as the comprehensive cost of the path trajectory from the starting point to the current node. It is expressed as the cost from time t to time t+1 to the current node. When the trajectory point rectangle in the second predicted value map at time t passes through or reaches a grid greater than the upper limit value H of the region, the node enters the closed set, otherwise it enters the open set and adds the t+1 time attribute to the node, and introduces pheromone to accelerate the search. h(t+1) is the estimated cost from a node at time t to the target node, expressed as x, y, v and The total cost, for example

The technical solution of the embodiment of the present application is adopted, and the first value map is generated by hierarchically distinguishing the drivable areas through high-precision map elements, predicting the target trajectory and reusing the first value map, generating the second value map time series, determining the drivable area threshold according to the decision signal, reusing the second value map time series multiple times, modifying the search space and sampling space and cost function in A*, and adding pheromones at the same time, generating fast, low-computing power and comfortable and smooth trajectories. Since the lane topology constraints, traffic rules constraints and abnormal situation constraints are added to the map, the technical solution of the embodiment of the present application can be applied to more situations and is more in line with the actual situation; the amount of calculation is reduced by reusing the first value map; the search speed is improved by adding pheromones; by converting the sampling space into the unit time speed and direction angle change, trajectory points are added between nodes, so that the optimized trajectory is smoother and the path optimization effect is improved.

All the above optional technical solutions can be arbitrarily combined to form optional embodiments of the present application, which will not be described one by one here.

The following are device embodiments of the present application, which can be used to execute the method embodiments of the present application. For details not disclosed in the device embodiments of the present application, please refer to the method embodiments of the present application.

FIG8 is a schematic diagram of a path optimization device provided in an embodiment of the present application. As shown in FIG8 , the device includes:

The acquisition module 801 is configured to acquire the navigation route map and obstacle information of the vehicle's current travel, where the obstacle is an obstacle in the navigation route. The obstacle information includes at least the location information and attribute information of static obstacles, and the predicted trajectory of dynamic obstacles.

The first map generation module 802 is configured to perform rasterization processing on the navigation route map, and assign values to each grid based on the location information and attribute information of the static obstacles to obtain a first-time predicted value map.

The second map generation module 803 is configured to update the first-time predicted value map based on the predicted trajectory of the dynamic obstacle to obtain a second-time predicted value map, wherein the first-time predicted value map and the second-time predicted value map are used to indicate the distance between the vehicle and the obstacle at different times.

The determination module 804 is configured to determine a drivable area of the vehicle from time t to time t+1 based on the second time prediction value map.

The planning module 805 is configured to obtain the state information of the vehicle at time t and the target search step length, and perform local path planning based on the state information of the vehicle at time t and the target search step length according to the dynamic constraints of the Ackerman chassis model to obtain a set of predicted trajectories of the vehicle from time t to time t+1, where t is greater than 0 and less than the total driving time of the vehicle.

The optimization module 806 is configured to determine a target trajectory from the set of predicted trajectories based on the second time prediction value map, and use the target trajectory as an optimized path of the vehicle from time t to time t+1.

According to the technical solution provided in the embodiment of the present application, by obtaining the navigation route map and obstacle information of the vehicle's current travel, the navigation route map is rasterized based on the static obstacle information in the obstacle information to obtain a first-time predicted value map, and the first-time predicted value map is updated based on the dynamic obstacle information in the obstacle information to obtain a second-time predicted value map, and a local path planning algorithm is used to obtain a set of predicted trajectories of the vehicle from time t to time t+1 that satisfy the dynamic constraints of the Ackerman chassis model, and the second-time predicted value map is used to determine the target trajectory from the predicted trajectory set, and the target trajectory is used as the optimized path of the vehicle from time t to time t+1, thereby enriching the constraints during local path planning, and a smooth trajectory can be obtained through one calculation, thereby improving the efficiency of path optimization, and the optimized path reduces the possibility of swinging back and forth, thereby improving the driving comfort of the vehicle and enhancing the user experience.

In an embodiment of the present application, the attribute information of the static obstacle includes at least: type information of the static obstacle and passage condition information of the static obstacle; in the first time predicted value map, the grid value is negatively correlated with the allowed passage degree of the static obstacle included in the grid; wherein the allowed passage degree of the static obstacle included in each grid is determined in the following manner: in response to determining that the static obstacle is a collision type obstacle, determining the allowed passage degree of the static obstacle to be a first value; in response to determining that the static obstacle is an obstacle that allows temporary passage, determining the allowed passage degree of the static obstacle to be a second value; in response to determining that the static obstacle is an obstacle that allows conditional passage, determining the allowed passage degree of the static obstacle to be a third value; in response to determining that the static obstacle is an obstacle that does not need to be occupied during normal driving, determining the allowed passage degree of the static obstacle to be a fourth value; in response to determining that the static obstacle is an obstacle that allows passage, determining the allowed passage degree of the static obstacle to be a fifth value; wherein the first value is less than the second value, the second value is less than the third value, the third value is less than the fourth value, and the fourth value is less than the fifth value.

In an embodiment of the present application, the second time prediction value map is a time series map; the first time prediction value map is updated based on the predicted trajectory of the dynamic obstacle to obtain the second time prediction value map, including: obtaining the target prediction position of the dynamic obstacle at time t; calculating the possible collision area of the dynamic obstacle at time t based on the speed of the dynamic obstacle at time t, the road condition information at time t, and the weather information at time t; in the first time prediction value map, the grid value at the target prediction position is updated to the maximum value, and the grid value in the possible collision area is updated to the second largest value to obtain the second time prediction value map.

In an embodiment of the present application, before determining the vehicle's drivable area, the method also includes: setting an area upper limit value, the area upper limit value is used to represent the maximum value of each grid in the vehicle's drivable area; setting an open set and a closed set, the open set is the nodes corresponding to the grids that may be passed during path planning, and the closed set is the nodes corresponding to the grids that will not be passed during path planning and the grids that have been added to the optimized path; determining the vehicle's drivable area from time t to time t+1 based on the second time predicted value map, including: determining the target open set of the vehicle from time t to time t+1 according to the second time predicted value map at time t and time t+1; determining the area composed of grids whose grid values corresponding to the nodes in the target open set are less than the area upper limit value, which is the drivable area of the vehicle from time t to time t+1.

In the embodiment of the present application, the state information of the vehicle at time t is: Where _xt is the horizontal coordinate value of the vehicle at time t, _yt is the vertical coordinate value of the vehicle at time t, and _vt is the speed of the vehicle at time t. is the heading angle of the vehicle at time t, _θt is the tire angle of the vehicle at time t; the target search step is Δt, Δt=ΔT/N, ΔT is the preset search step, N represents the number of samples from time t to time t+1, and N is a positive integer greater than 1; the dynamic constraints of the Ackerman chassis model include constraints on vehicle speed and vehicle tire angle.

In an embodiment of the present application, local path planning is performed in a drivable area based on the state information of the vehicle at time t and the target search step length, including: based on the state information of the vehicle at time t, M sampling combinations (Δv _t,i , Δθ _t,i ) are determined, where i is a positive integer from 1 to M, and M is a positive integer greater than 1; for each sampling combination, a path planning algorithm is used to predict M predicted trajectory sets of the vehicle at time t+1 that satisfy the dynamic constraints of the Ackerman chassis model in the drivable area according to the target search step length; wherein each predicted trajectory set includes N predicted position nodes, each predicted position node corresponds to a search step length, each predicted position node is obtained by searching the grid sequence number corresponding to the node, and the vehicle state information corresponding to each predicted position node is x _t+1 is the horizontal coordinate value of the vehicle at time t+1, y _t+1 is the vertical coordinate value of the vehicle at time t+1, and v _t+1 is the speed of the vehicle at time t+1. is the heading angle of the vehicle at time t+1, and θ _t+1 is the tire angle of the vehicle at time t+1.

In the embodiment of the present application, the vehicle status information corresponding to each predicted location node Determined in the following way: Wherein, n is a positive integer greater than 0 and less than N.

In an embodiment of the present application, a target trajectory is determined from a set of predicted trajectories based on a second time predicted value map, including: generating an evaluation function based on the second time predicted value map; using the evaluation function to calculate a cost value of each predicted position node in the set of predicted trajectories; determining the sum of the cost values of each predicted position node in each predicted trajectory as the cost value of each predicted trajectory; and determining the predicted trajectory with the smallest cost value as the target trajectory.

In an embodiment of the present application, the evaluation function includes a first evaluation item and a second evaluation item; the first evaluation item is the sum of the cost of the vehicle from the position of the vehicle at the starting moment to the position of the vehicle at time t, and the cost of the vehicle from the position of the vehicle at time t to the predicted position node to be evaluated; the second evaluation item is the estimated cost of the vehicle from the predicted position node to be evaluated to the target node, and the target node is the end position in the predicted trajectory where the predicted position node to be evaluated is located; the evaluation function is used to calculate the cost of each predicted position node in the predicted trajectory set, including: using the evaluation function to calculate the first evaluation item value of the predicted position node to be evaluated; using the evaluation function to calculate the second evaluation item value of the predicted position node to be evaluated; using the sum of the first evaluation item value and the second evaluation item value as the cost value of the predicted position node to be evaluated, and the predicted position node to be evaluated is any node among the predicted position nodes in the predicted trajectory set.

In an embodiment of the present application, after using the evaluation function to calculate the cost value of each predicted position node in the predicted trajectory set that is located in the open set, the method also includes: in response to the first evaluation item of the predicted position node to be evaluated being greater than or equal to the regional upper limit value, adding the predicted position node to be evaluated to the closed set, discarding the predicted trajectory where the predicted position node to be evaluated is located, or regenerating the predicted trajectory where the predicted position node to be evaluated is located; in response to the first evaluation item of the predicted position node to be evaluated being less than the regional upper limit value, adding the predicted position node to be evaluated to the open set, and adding a time attribute of time t+1 in the predicted position node to be evaluated.

In an embodiment of the present application, in the open set, each node contains pheromone; after the predicted position node to be evaluated is added to the closed set, the method also includes: determining the target pheromone of the predicted position node to be evaluated added to the closed set; subtracting the target pheromone from the pheromone of each node in the open set whose distance to the predicted position node to be evaluated added to the closed set is less than a preset distance threshold, and adding the target pheromone to the node corresponding to the vehicle's position at time t.

In an embodiment of the present application, the second evaluation item value is obtained by at least weighted summing up the distance estimated cost, speed estimated cost and heading estimated cost of the vehicle from the predicted position node to be evaluated to the target node.

It should be understood that the size of the serial numbers of the steps in the above embodiments does not mean the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of the present application.

FIG9 is a schematic diagram of an electronic device provided in an embodiment of the present application. As shown in FIG9 , the electronic device 9 of this embodiment includes: a processor 901, a memory 902, and a computer program 903 stored in the memory 902 and executable on the processor 901. When the processor 901 executes the computer program 903, the steps in the above-mentioned various method embodiments are implemented. Alternatively, when the processor 901 executes the computer program 903, the functions of each module/unit in the above-mentioned various device embodiments are implemented.

The electronic device 9 may be a desktop computer, a notebook, a PDA, a cloud server, or other electronic device. The electronic device 9 may include, but is not limited to, a processor 901 and a memory 902. Those skilled in the art will appreciate that FIG. 9 is merely an example of the electronic device 9 and does not limit the electronic device 9, and may include more or fewer components than shown in the figure, or different components.

The processor 901 may be a central processing unit (CPU), or other general-purpose processors, digital signal processors (DSP), application-specific integrated circuits (ASIC), field-programmable gate arrays (FPGA), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc.

The memory 902 may be an internal storage unit of the electronic device 9, for example, a hard disk or memory of the electronic device 9. The memory 902 may also be an external storage device of the electronic device 9, for example, a plug-in hard disk, a smart media card (SMC), a secure digital (SD) card, a flash card, etc. equipped on the electronic device 9. The memory 902 may also include both an internal storage unit of the electronic device 9 and an external storage device. The memory 902 is used to store computer programs and other programs and data required by the electronic device.

Those skilled in the art can clearly understand that for the convenience and simplicity of description, only the division of the above-mentioned functional units and modules is used as an example for illustration. In actual applications, the above-mentioned functions can be distributed and completed by different functional units and modules as needed, that is, the internal structure of the device is divided into different functional units or modules to complete all or part of the functions described above. The functional units and modules in the embodiments can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit. The above-mentioned integrated unit can be implemented in the form of hardware or in the form of software functional units.

If the integrated module/unit is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the present application implements all or part of the processes in the above-mentioned embodiment method, and can also be completed by instructing the relevant hardware through a computer program. The computer program can be stored in a computer-readable storage medium, and the computer program can implement the steps of the above-mentioned various method embodiments when executed by the processor. The computer program may include computer program code, which may be in source code form, object code form, executable file or some intermediate form. Computer-readable media may include: any entity or device capable of carrying computer program code, recording medium, U disk, mobile hard disk, disk, optical disk, computer memory, read-only memory (ROM), random access memory (RAM), electric carrier signal, telecommunication signal and software distribution medium, etc.

The above embodiments are only used to illustrate the technical solutions of the present application, rather than to limit them. Although the present application has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or make equivalent replacements for some of the technical features therein. These modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the embodiments of the present application, and should all be included in the protection scope of the present application.

Claims (14)

1. A method for path optimization, the method for optimizing a vehicle navigation path, the method comprising:

Acquiring a navigation route map and obstacle information of the current running of the vehicle, wherein the obstacle is an obstacle in the navigation route, and the obstacle information at least comprises position information and attribute information of a static obstacle and a predicted track of a dynamic obstacle;

Performing rasterization processing on the navigation route map, and assigning a value to each grid based on the position information and the attribute information of the static obstacle to obtain a first time prediction value map;

Updating the first time prediction value map based on the prediction track of the dynamic obstacle to obtain a second time prediction value map, wherein the first time prediction value map and the second time prediction value map are used for representing the distance degree between the vehicle and the obstacle at different moments;

determining a drivable area of the vehicle from time t to time t+1 based on the second time predictive value map;

Acquiring state information and a target search step length of a vehicle at a moment t, and carrying out local path planning in the drivable area based on the state information and the target search step length of the vehicle at the moment t to obtain a predicted track set of the vehicle conforming to the dynamic constraint of the Ackerman chassis model from the moment t to a moment t+1, wherein t is more than 0 and less than the total running time of the vehicle;

Determining a target track from the predicted track set based on the second time prediction value map, and taking the target track as an optimized path of the vehicle from time t to time t+1;

wherein, the attribute information of the static obstacle at least comprises: type information of static obstacles and traffic condition information of the static obstacles;

in the first time prediction value map, a grid value is inversely related to the allowable passing degree of static obstacles included in the grid;

The allowable passing degree of the static barriers included in each grid is determined by the following mode:

in response to determining that a static obstacle is a collision type obstacle, determining that the allowable passing degree of the static obstacle is a first value;

in response to determining that a static obstacle is an allowed temporary pass obstacle, determining that an allowed pass degree of the static obstacle is a second value;

In response to determining that a static obstacle is a conditional access allowable obstacle, determining that the level of access allowable for the static obstacle is a third value;

In response to determining that the static obstacle does not need to occupy the obstacle when the vehicle is running normally, determining that the allowable passing degree of the static obstacle is a fourth value;

In response to determining that a static obstacle is an allowed passage obstacle, determining that an allowed passage degree of the static obstacle is a fifth value;

The first value is smaller than the second value, the second value is smaller than the third value, the third value is smaller than the fourth value, and the fourth value is smaller than the fifth value.

2. The method of claim 1, wherein the second time-predictive value map is a time-series map;

The updating the first time prediction value map based on the prediction track of the dynamic obstacle to obtain a second time prediction value map comprises the following steps:

acquiring a target predicted position of the dynamic obstacle at a time t;

Calculating a possible collision area of the dynamic obstacle at the moment t based on the speed of the dynamic obstacle at the moment t, the road condition information of the moment t and the weather information of the moment t;

And in the first time prediction value map, updating the grid value at the target prediction position to be the maximum value, and updating the grid value in the possible collision area to be the next-largest value, so as to obtain the second time prediction value map.

3. The method of claim 1, wherein prior to determining the drivable region of the vehicle, the method further comprises:

Setting an area upper limit value which is used for representing the maximum value of each grid in the vehicle running area;

Setting an open set and a closed set, wherein the open set is a node corresponding to a grid which can possibly pass through during path planning, and the closed set is a node corresponding to a grid which can not pass through during path planning and a grid which is added with an optimized path;

The determining the drivable area of the vehicle from the time t to the time t+1 based on the second time prediction value map comprises:

determining a target open set of the vehicle from the time t to the time t+1 according to the second time prediction value map of the time t and the time t+1;

And determining an area formed by grids of which the grid values of the grids corresponding to the nodes are smaller than the upper limit value of the area in the target open set as a drivable area from the moment t to the moment t+1 of the vehicle.

4. The method of claim 3, wherein the vehicle's status information at time t is (x _t,y_t,v_t,，) Where x _t is the abscissa value of the vehicle at time t, y _t is the ordinate value of the vehicle at time t, v _t is the speed of the vehicle at time t,For the heading angle of the vehicle at time t,The tire included angle of the vehicle at the time t;

The target searching step length is deltat, deltat=deltat/N, deltat is a preset searching step length, N represents the sampling number from time T to time t+1, and N is a positive integer greater than 1;

The dynamic constraints of the ackerman chassis model include constraints on vehicle speed and vehicle tire angle.

5. The method of claim 4, wherein the performing local path planning in the drivable region based on the state information of the vehicle at time t and the target search step comprises:

based on the state information of the vehicle at the time t, M sampling combinations are determined ，) I is a positive integer from 1 to M, M is a positive integer greater than 1;

Predicting each sampling combination by using a path planning algorithm according to the target search step length in the drivable area to obtain M predicted track sets of the vehicle at a time t+1, wherein the M predicted track sets meet the dynamics constraint of an Ackerman chassis model;

wherein each predicted track set comprises N predicted position nodes, each predicted position node corresponds to one search step length, each predicted position node is obtained through searching of a grid serial number corresponding to the node, vehicle state information corresponding to each predicted position node is (x _t+1,y_t+1,v_t+1,φ_t+1,θ_t+1）,x_t+1 is an abscissa value of a vehicle at a time t+1, y _t+1 is an ordinate value of the vehicle at the time t+1, v _t+1 is a speed of the vehicle at the time t+1, For the heading angle of the vehicle at time t +1,Is the tire angle of the vehicle at time t+1.

6. The method according to claim 5, wherein the vehicle state information (x _t+1,y_t+1,v_t+1,φ_t+1,θ_t+1) for each predicted position node is determined as follows:

Wherein N is a positive integer greater than 0 and less than N, j is a positive integer greater than or equal to 0 and less than or equal to N, and l is the wheelbase of the front and rear wheels.

7. The method of claim 5 or 6, wherein the determining a target trajectory from the set of predicted trajectories based on the second time predictive value map comprises:

generating an evaluation function based on the second time predictive value map;

calculating a cost value of each predicted position node in the predicted track set by using the evaluation function;

Determining the sum of the cost values of the prediction position nodes in each prediction track as the cost value of each prediction track;

and determining the predicted track with the minimum cost value as the target track.

8. The method of claim 7, wherein the evaluation function comprises a first evaluation term and a second evaluation term;

the first evaluation item is the sum of the cost value from the position of the vehicle at the starting moment to the position of the vehicle at the moment t and the cost value from the position of the vehicle at the moment t to the node of the predicted position to be evaluated;

the second evaluation item is the estimated cost of the vehicle from the predicted position node to be evaluated to a target node, and the target node is the end position of the predicted track where the predicted position node to be evaluated is located;

the calculating a cost value of each predicted position node in the predicted track set by using the evaluation function includes:

calculating a first evaluation item value of the predicted position node to be evaluated by using the evaluation function;

calculating a second evaluation item value of the predicted position node to be evaluated by using the evaluation function;

and taking the sum of the first evaluation item value and the second evaluation item value as a cost value of the predicted position node to be evaluated, wherein the predicted position node to be evaluated is any node in all the predicted position nodes in the predicted track set.

9. The method of claim 8, wherein after calculating the cost value for each predicted location node in the set of predicted trajectories that is in an open set using the evaluation function, the method further comprises:

Responding to the fact that a first evaluation item of the predicted position node to be evaluated is larger than or equal to the upper limit value of the area, adding the predicted position node to be evaluated into a closed set, discarding a predicted track where the predicted position node to be evaluated is located, or regenerating the predicted track where the predicted position node to be evaluated is located;

And adding the predicted position node to be evaluated into an open set in response to the first evaluation item of the predicted position node to be evaluated being smaller than the upper limit value of the area, and adding the time attribute of the time t+1 into the predicted position node to be evaluated.

10. The method of claim 9, wherein each node in the open set comprises a pheromone;

after adding the predicted location node to be evaluated to the closed set, the method further comprises:

Determining target pheromones of the to-be-evaluated predicted position nodes added into the closed set;

Subtracting the target pheromone from the pheromone of each node, which is in the open set and has a distance smaller than a preset distance threshold, of the nodes of the predicted position to be evaluated added into the closed set, and adding the target pheromone to the node corresponding to the position of the vehicle at the moment t.

11. The method of claim 8, wherein the second evaluation term value is derived from at least a weighted sum of a distance estimated cost, a speed estimated cost, and a heading estimated cost of the vehicle from the predicted location node to be evaluated to a target node.

12. A path optimization apparatus, comprising:

The system comprises an acquisition module, a prediction module and a control module, wherein the acquisition module is configured to acquire a navigation route map of the current running of a vehicle and obstacle information, wherein the obstacle is an obstacle in the navigation route, and the obstacle information at least comprises position information and attribute information of a static obstacle and a prediction track of a dynamic obstacle;

the first map generation module is configured to perform rasterization processing on the navigation route map, and assign a value to each grid based on the position information and the attribute information of the static obstacle to obtain a first time prediction value map;

A second map generation module configured to update the first time prediction value map based on the predicted track of the dynamic obstacle to obtain a second time prediction value map, wherein the first time prediction value map and the second time prediction value map are used for representing the distance degree between the vehicle and the obstacle at different moments;

A determination module configured to determine a drivable region of the vehicle from a time t to a time t+1 based on the second time predictive value map;

The planning module is configured to acquire state information of a vehicle at a time t and a target searching step length, and based on the state information of the vehicle at the time t and the target searching step length, the planning module performs local path planning according to dynamic constraints of an Ackerman chassis model to obtain a predicted track set of the vehicle from the time t to a time t+1, wherein t is greater than 0 and less than the total running time of the vehicle;

an optimization module configured to determine a target track from the predicted track set based on the second time prediction value map, and take the target track as an optimized path of the vehicle from time t to time t+1;

wherein, the attribute information of the static obstacle at least comprises: type information of static obstacles and traffic condition information of the static obstacles;

in the first time prediction value map, a grid value is inversely related to the allowable passing degree of static obstacles included in the grid;

The allowable passing degree of the static barriers included in each grid is determined by the following mode:

in response to determining that a static obstacle is a collision type obstacle, determining that the allowable passing degree of the static obstacle is a first value;

in response to determining that a static obstacle is an allowed temporary pass obstacle, determining that an allowed pass degree of the static obstacle is a second value;

In response to determining that a static obstacle is a conditional access allowable obstacle, determining that the level of access allowable for the static obstacle is a third value;

In response to determining that the static obstacle does not need to occupy the obstacle when the vehicle is running normally, determining that the allowable passing degree of the static obstacle is a fourth value;

In response to determining that a static obstacle is an allowed passage obstacle, determining that an allowed passage degree of the static obstacle is a fifth value;

The first value is smaller than the second value, the second value is smaller than the third value, the third value is smaller than the fourth value, and the fourth value is smaller than the fifth value.

13. An electronic device comprising a memory, a processor and a computer program stored in the memory and executable on the processor, characterized in that the processor implements the steps of the method according to any of claims 1 to 11 when the computer program is executed.

14. A computer readable storage medium storing a computer program, characterized in that the computer program when executed by a processor implements the steps of the method according to any one of claims 1 to 11.