A method and device for processing vehicle lane change

Abstract

The present invention discloses a vehicle lane change processing method and device, which relates to the field of autonomous driving technology. A specific implementation of the method includes: receiving the current driving environment parameters transmitted by the target vehicle to determine the information of other vehicles located around the target vehicle; taking the target vehicle as the starting point, generating multiple traffic orders by selecting vehicles and corresponding vehicle information from other vehicles; screening out the target traffic order from the multiple traffic orders, and planning conflict-free trajectories for each vehicle in turn according to the vehicle priority order in the target traffic order, so that the target vehicle and other vehicles do not collide when driving at the same time; triggering the target vehicle to perform a lane change operation according to the conflict-free trajectory planned for the target vehicle. This implementation adopts the Monte Carlo tree search method and heuristic rules, which can quickly screen out the target traffic order, match the corresponding trajectory from the preset trajectory set according to the driving parameters, improve the efficiency of lane change decision generation, and ensure lane change safety.

Description

Technical Field

[0001] The present invention relates to the field of autonomous driving technology, and in particular to a vehicle lane change processing method and device. Background Art

[0002] Lane changing strategy is a key technology to ensure the driving safety of autonomous vehicles. However, it is still difficult to make the vehicle make safe and satisfactory lane changing decisions when driving in complex dynamic environments. Although many lane changing decision methods have been proposed, most of them are optimal planning strategies. They are solved by designing solution algorithms or directly using commercial software. The solution time is usually long, and even due to nonlinear constraints, it is impossible to obtain a feasible solution.

[0003] As an optimization method of the above implementation, a feasible planning strategy is proposed to determine the passage order according to the heuristic rules designed by humans, and then optimize the trajectories of all vehicles according to the obtained passage order. However, the passage order given by the heuristic rules mostly obeys the first-in-first-out principle, which is difficult to produce good traffic performance. Summary of the invention

[0004] In view of this, an embodiment of the present invention provides a vehicle lane change processing method and device for handling conflicts at different levels, while taking into account both traffic coordination performance and computing time, so as to at least solve the problem that it is difficult to determine the lane change strategy of an autonomous driving vehicle in the prior art.

[0005] To achieve the above object, according to one aspect of an embodiment of the present invention, a vehicle lane change processing method is provided, comprising:

[0006] Receiving current driving environment parameters transmitted by the target vehicle to determine information of other vehicles located around the target vehicle;

[0007] Taking the target vehicle as a starting point, generating a plurality of passing orders by selecting vehicles and corresponding vehicle information from the other vehicles;

[0008] Filtering a target passing order from the multiple passing orders, and planning a conflict-free trajectory for each vehicle in turn according to the vehicle priority order in the target passing order, so that the target vehicle and the other vehicles do not collide when traveling at the same time;

[0009] According to the conflict-free trajectory planned for the target vehicle, the target vehicle is triggered to perform a lane change operation.

[0010] Optionally, before determining the information of other vehicles located around the target vehicle, the method further includes:

[0011] If, in the current driving environment parameters, the road resources ahead of the lane where the target vehicle is currently traveling are occupied or the road geometry changes, then the lane change logic is triggered; or

[0012] If the current driving environment parameters include a lane change request transmitted by the target vehicle, the lane change logic is triggered.

[0013] Optionally, also include:

[0014] When receiving current driving environment parameters transmitted by multiple vehicles, priority is given to allocating road rights to vehicles that are in a critical conflict zone ahead; wherein the critical conflict zone is an area where road resources are occupied or road geometry has changed; and

[0015] When there is no critical conflict zone ahead of each vehicle, the right of way is allocated preferentially to the vehicle at the front among the multiple vehicles.

[0016] Optionally, the determining the information of other vehicles located around the target vehicle further includes:

[0017] Based on the current areas where the target vehicle and the other vehicles are located, a local conflict area is constructed, and an available action of each other vehicle in the local conflict area is determined.

[0018] Optionally, the generating of multiple passage sequences by selecting vehicles and corresponding vehicle information from the other vehicles with the target vehicle as the starting point includes:

[0019] Taking the target vehicle as a starting point, a plurality of passage sequences are generated by randomly selecting vehicles and the movable actions of each vehicle from among the other vehicles.

[0020] Optionally, the filtering out a target pass sequence from the multiple pass sequences includes:

[0021] Invoke a heuristic rule to select one or more passing sequences from the plurality of passing sequences according to a first-in-first-out principle based on the current positions of the target vehicle and each other vehicle;

[0022] Calling a Monte Carlo tree search program to determine a node with an unexpanded action in each selected passage sequence, and creating a new node as a child node of the node; wherein the child node is used to indicate a driving environment parameter of the target vehicle after executing the unexpanded action;

[0023] Taking the child node as the starting point, sampling is performed from vehicles that are not included in each of the passing sequences, the sampled vehicles are added to each of the passing sequences, and the above random sampling operation is repeated until a leaf node is generated, thereby obtaining a simulation score;

[0024] The nodes in each passage sequence and all nodes on the path from the root node to the nodes are scored cumulatively according to the simulation score; wherein the root node is used to indicate the current driving environment parameters of the target vehicle;

[0025] The confidence value is calculated according to the number of simulations and the cumulative score of each node, the target node with the largest confidence value is determined, and the passage order between the target node and the root node is used as the target passage order.

[0026] Optionally, before planning a conflict-free trajectory for each vehicle in sequence according to the vehicle priority order in the target traffic order, the method further includes:

[0027] Calculating a safe distance between the target vehicle and the vehicle in front of the lane-changing target lane so that the target vehicle does not collide with the vehicle in front of the lane-changing target lane according to the driving parameters of the target vehicle and the driving parameters of the vehicle in front of the lane-changing target lane;

[0028] In response to a distance between the target vehicle and a preceding vehicle in the lane-changing target lane being less than a safe vehicle distance, adjusting a speed of the target vehicle so that the distance between the target vehicle and a preceding vehicle in the lane-changing target lane is greater than or equal to the safe vehicle distance;

[0029] The above operation of adjusting the speed to meet the collision constraint condition is repeated so that the distance between the target vehicle and each other vehicle and the corresponding preceding vehicle all meets the collision constraint condition.

[0030] Optionally, planning a conflict-free trajectory for each vehicle in turn includes:

[0031] According to the driving parameters of each vehicle, searching for a conflict-free trajectory matching the driving parameters from a preset trajectory set; wherein the driving parameters include speed, acceleration, and vehicle type;

[0032] When there are multiple matched conflict-free trajectories, the conflict-free trajectory with the highest right-of-way priority among the multiple conflict-free trajectories is used as the target conflict-free trajectory.

[0033] To achieve the above object, according to another aspect of an embodiment of the present invention, a vehicle lane change processing device is provided, comprising:

[0034] A determination module, used for receiving current driving environment parameters transmitted by a target vehicle to determine information of other vehicles located around the target vehicle;

[0035] A generating module, configured to generate a plurality of passing sequences by selecting vehicles and corresponding vehicle information from the other vehicles, taking the target vehicle as a starting point;

[0036] A processing module, configured to select a target passing sequence from the plurality of passing sequences, and plan a conflict-free trajectory for each vehicle in turn according to the vehicle priority order in the target passing sequence, so that the target vehicle and the other vehicles do not collide when traveling at the same time;

[0037] The lane changing module is used to trigger the target vehicle to perform a lane changing operation according to the conflict-free trajectory planned for the target vehicle.

[0038] Optionally, a lane change judgment module is also included, which is used to:

[0039] If, in the current driving environment parameters, the road resources ahead of the lane where the target vehicle is currently traveling are occupied or the road geometry changes, then the lane change logic is triggered; or

[0040] If the current driving environment parameters include a lane change request transmitted by the target vehicle, the lane change logic is triggered.

[0041] Optionally, a lane change priority determination module is also included, which is used to:

[0042] When receiving current driving environment parameters transmitted by multiple vehicles, priority is given to allocating road rights to vehicles that are in a critical conflict zone ahead; wherein the critical conflict zone is an area where road resources are occupied or road geometry has changed; and

[0043] When there is no critical conflict zone ahead of each vehicle, the right of way is allocated preferentially to the vehicle at the front among the multiple vehicles.

[0044] Optionally, the determining module is further used to:

[0045] Based on the current areas where the target vehicle and the other vehicles are located, a local conflict area is constructed, and an available action of each other vehicle in the local conflict area is determined.

[0046] Optionally, the generation module is used to: take the target vehicle as the starting point, and generate multiple passage sequences by randomly selecting vehicles and the movable actions of each vehicle from the other vehicles.

[0047] Optionally, the processing module is used to:

[0048] Invoke a heuristic rule to select one or more passing sequences from the plurality of passing sequences according to a first-in-first-out principle based on the current positions of the target vehicle and each other vehicle;

[0049] Calling a Monte Carlo tree search program to determine a node with an unexpanded action in each selected passage sequence, and creating a new node as a child node of the node; wherein the child node is used to indicate a driving environment parameter of the target vehicle after executing the unexpanded action;

[0050] Taking the child node as the starting point, sampling is performed from vehicles that are not included in each of the passing sequences, the sampled vehicles are added to each of the passing sequences, and the above random sampling operation is repeated until a leaf node is generated, thereby obtaining a simulation score;

[0051] The nodes in each passage sequence and all nodes on the path from the root node to the nodes are scored cumulatively according to the simulation score; wherein the root node is used to indicate the current driving environment parameters of the target vehicle;

[0052] The confidence value is calculated according to the number of simulations and the cumulative score of each node, the target node with the largest confidence value is determined, and the passage order between the target node and the root node is used as the target passage order.

[0053] Optionally, the processing module is further used to:

[0054] Calculating a safe distance between the target vehicle and the vehicle in front of the lane-changing target lane so that the target vehicle does not collide with the vehicle in front of the lane-changing target lane according to the driving parameters of the target vehicle and the driving parameters of the vehicle in front of the lane-changing target lane;

[0055] In response to a distance between the target vehicle and a preceding vehicle in the lane-changing target lane being less than a safe vehicle distance, adjusting a speed of the target vehicle so that the distance between the target vehicle and a preceding vehicle in the lane-changing target lane is greater than or equal to the safe vehicle distance;

[0056] The above operation of adjusting the speed to meet the collision constraint condition is repeated so that the distance between the target vehicle and each other vehicle and the corresponding preceding vehicle all meets the collision constraint condition.

[0057] Optionally, the processing module is used to:

[0058] According to the driving parameters of each vehicle, searching for a conflict-free trajectory matching the driving parameters from a preset trajectory set; wherein the driving parameters include speed, acceleration, and vehicle type;

[0059] When there are multiple matched conflict-free trajectories, the conflict-free trajectory with the highest right-of-way priority among the multiple conflict-free trajectories is used as the target conflict-free trajectory.

[0060] To achieve the above objective, according to another aspect of an embodiment of the present invention, a vehicle lane-changing decision electronic device is provided.

[0061] The electronic device of an embodiment of the present invention includes: one or more processors; a storage device for storing one or more programs, when the one or more programs are executed by the one or more processors, the one or more processors implement any of the above-mentioned vehicle lane change processing methods.

[0062] To achieve the above objective, according to another aspect of an embodiment of the present invention, a computer-readable medium is provided, on which a computer program is stored, and when the program is executed by a processor, any of the above-mentioned vehicle lane change processing methods is implemented.

[0063] According to the solution provided by the present invention, an embodiment of the above invention has the following advantages or beneficial effects: fully considering the complexity and dynamics of the road, evaluating the necessity of lane changing for the target vehicle, and making preparations for lane changing in advance; fully considering the feasible actions of other vehicles around the target vehicle, generating multiple traffic orders, relying on the Monte Carlo tree search method and heuristic rules to accurately determine the target traffic order; when all vehicles meet the collision constraints, planning a conflict-free path to ensure the safety of the target vehicle's lane changing process.

[0064] The further effects of the above-mentioned non-conventional optional manner will be described below in conjunction with the specific implementation manner. BRIEF DESCRIPTION OF THE DRAWINGS

[0065] The accompanying drawings are used to better understand the present invention and do not constitute an improper limitation of the present invention.

[0066] Figure 1 is a schematic diagram of the main flow of a vehicle lane change processing method according to an embodiment of the present invention;

[0067] Figure 2(a) is a schematic diagram of a typical traffic scene with lane change;

[0068] Figure 2(b) is a schematic diagram of another typical traffic scenario with lane change;

[0069] Figure 2(c) is a schematic diagram of finding multiple matching non-conflicting trajectories from a trajectory set;

[0070] Figure 3 is a flow chart of an optional vehicle lane change processing method according to an embodiment of the present invention;

[0071] Figure 4 is a schematic diagram of search iterations using the Monte Carlo tree search method;

[0072] Figure 5 A schematic flow chart of another optional vehicle lane change processing method according to an embodiment of the present invention;

[0073] Figure 6is a schematic diagram of adjusting the vehicle speed to meet the collision constraint condition;

[0074] Figure 7 is a schematic diagram of main modules of a vehicle lane change processing device according to an embodiment of the present invention;

[0075] Figure 8 is an exemplary system architecture diagram to which embodiments of the present invention may be applied;

[0076] Fig. 9 It is a schematic diagram of the structure of a computer system of a mobile device or a server suitable for implementing an embodiment of the present invention. DETAILED DESCRIPTION

[0077] The following is a description of exemplary embodiments of the present invention in conjunction with the accompanying drawings, including various details of the embodiments of the present invention to facilitate understanding, which should be considered as merely exemplary. Therefore, it should be recognized by those of ordinary skill in the art that various changes and modifications may be made to the embodiments described herein without departing from the scope and spirit of the present invention. Similarly, for clarity and conciseness, the description of well-known functions and structures is omitted in the following description.

[0078] It should be noted that the embodiments and features of the embodiments of the present invention can be combined with each other without conflict. The acquisition, storage, use, and processing of data in the technical solution of this application are in compliance with the relevant provisions of national laws and regulations.

[0079] The terms involved in this plan are explained as follows:

[0080] Heuristic rules: Heuristics is the combination of simplified virtual machines and simplified behavior judgment engines. Based on heuristic rules, the focus is on updating the feature value recognition technology and solving the defects of single feature code comparison. The purpose is not to detect all unknown viruses, but only to supplement the feature value scanning technology.

[0081] Monte Carlo tree search (MCTS) is a Monte Carlo method based on tree structure. N (N is equal to the number of decisions, i.e. the tree depth) space to perform a heuristic search to find the optimal tree structure path (feasible solution) based on certain feedback. In general, MCTS is a heuristic random search algorithm driven by deterministic rules.

[0082] Traffic order: indicates the priority of each vehicle. Vehicles with higher rankings have the right to occupy road resources first.

[0083] See also Figure 1, which shows a main flow chart of a vehicle lane change processing method provided by an embodiment of the present invention, comprising the following steps:

[0084] S101: receiving current driving environment parameters transmitted by a target vehicle to determine information of other vehicles located around the target vehicle;

[0085] S102: Taking the target vehicle as a starting point, generating a plurality of passing sequences by selecting vehicles and corresponding vehicle information from the other vehicles;

[0086] S103: Filtering out a target passing sequence from the multiple passing sequences, and planning a conflict-free trajectory for each vehicle in turn according to the vehicle priority order in the target passing sequence, so that the target vehicle and the other vehicles do not collide when traveling at the same time;

[0087] S104: triggering the target vehicle to perform a lane change operation according to the conflict-free trajectory planned for the target vehicle.

[0088] In the above implementation, for step S101, intelligent driving research has currently become the forefront direction of development in the automotive field. In the future, as the number of vehicles on the road continues to increase, traffic congestion and traffic accidents will become more prominent. These problems are often caused by incorrect lane changes. Therefore, for autonomous driving vehicles, choosing a reasonable lane changing strategy is of great research value to ensure safe driving.

[0089] During the driving process of each vehicle, sensors are used to obtain real-time driving environment parameters, perceive the surrounding environment, and identify traffic scenes. Scenarios that require vehicles to change lanes can be divided into free lane changes and forced lane changes: 1) Forced lane changes are usually caused by traffic rules and road conditions. Vehicles must perform lane changes and lane change obstacle avoidance behaviors according to traffic rules. For example, when the sensor detects an accident ahead, a forced lane change is required. 2) Free lane changes are when the vehicle can autonomously choose to keep the lane or change lanes during driving.

[0090] This solution sets up conflict zones, which are divided into key conflict zones and local conflict zones according to their impact on traffic scenarios. When a vehicle is about to reach a key conflict zone, it is forced to change lanes. The key conflict zone is a relatively fixed area, usually set according to the characteristics of the road. It is a bottleneck area for traffic flow and hinders traffic flow. Common key conflict zones include urban intersections, highway ramps, and work areas. Therefore, when vehicles are driving in these areas, they need to carefully plan their driving trajectories to avoid collisions. There are two main reasons for key conflict zones:

[0091] 1. Road resources are occupied, such as vehicle accidents and work areas as shown in Figure 2(a). Part of the road area is disabled, resulting in a reduction in the number of lanes, affecting the traffic flow within a certain range, and even blocking the traffic flow of several roads. Therefore, all vehicles in the same lane as the accident vehicle or work area must change lanes in advance to avoid collisions. For example, in Figure 2(a), the front of the current lane of vehicle B is an impassable work area, so it needs to change lanes to the right in advance.

[0092] 2. Changes in road geometry, such as road intersections and road merging, lead to a sharp increase in collision risk and a decrease in traffic capacity. As shown in Figure 2(b), the key conflict area is the intersection, where the traffic and pedestrian flow are large, and the left-turning vehicle G needs to change lanes to the left lane due to traffic regulations.

[0093] The local conflict zone is opposite to the key conflict zone. It refers to the area where the vehicle action or behavior may cause conflict, and usually only affects the movement of several nearby vehicles within a short period of time. As shown in Figure 2(a), vehicle B needs to change lanes to the right. Vehicle C is already driving in the adjacent lane around vehicle B. Therefore, the local conflict zone is formed by the area where vehicles B and C are currently located, or the area where vehicles B, C and A are currently located. This area disappears after vehicle B completes the lane change. Vehicle E needs to change lanes to overtake. Vehicle D is already driving in the right lane around vehicle E. Therefore, the local conflict zone is formed by the area where vehicles E and D are currently located, or the area where vehicles C, D and E are currently located. This area disappears after vehicle E completes the lane change and overtaking. As shown in Figure 2(b), vehicle G needs to change lanes to the left. Vehicle F is already driving in the left lane around vehicle G. Therefore, the local conflict zone is formed by the area where vehicle F and G are currently located. This area disappears after vehicle G completes the lane change. That is, local conflict zones are generated by the action of a vehicle and disappear quickly with the end of the action.

[0094] From the above description, it can be seen that the critical conflict zone has a significant impact on the traffic efficiency of the entire traffic scenario. Therefore, in actual operation, it is necessary to first determine which vehicles are about to reach the critical conflict zone, and give priority to allocating road rights to vehicles that are about to reach the critical conflict zone (that is, they have the right to occupy road resources first). Then, the allocation result is fed back to the upstream to optimize the movement of the vehicle in the local conflict zone.

[0095] Furthermore, considering that the central controller on the roadside equipment is usually equipped with higher-performance computing devices than the on-board equipment, all vehicles in this solution communicate directly with the central controller and send the real-time driving environment parameters collected by their respective sensors to the central controller. The central controller then makes lane change decisions for each vehicle through its own high-performance computing devices or by accessing cloud computing.

[0096] It should be noted that processing the real-time driving environment parameters of each vehicle involves a balance between the flexibility and computing power of the computing system. In practical applications, this can be achieved through a self-organizing network. Specifically, the vehicle broadcasts its status and intention information, and all vehicles close to the conflict zone form a network through self-organization. After that, a leading vehicle is dynamically selected to plan a conflict-free trajectory for all vehicles, and the planned trajectory is returned to these vehicles. However, considering that the aforementioned central controller processing method can achieve faster calculation and faster implementation, and ensure the real-time collaborative decision-making and control of the vehicle group, this solution preferably adopts the aforementioned central controller processing method based on the needs of the scene and the evaluation of important factors.

[0097] The central controller receives the current driving environment parameters transmitted by the target vehicle, including the current driving lane, driving parameters, and surrounding scene information; the surrounding scene information includes: the distance between the target vehicle and the vehicle in front of the current lane, the distance between the target vehicle and the vehicle in front of the target lane for lane change, the distance between the target vehicle and the vehicle behind the target lane for lane change, the distance between the target vehicle and the vehicle behind the current lane, the use of the turn signal of the vehicle behind the current lane at the current moment, and driving parameters include the speed of the target vehicle at the current moment, the speed of the vehicle in front of the current lane, the speed of the vehicle in front of the target lane for lane change, the speed of the vehicle behind the target lane for lane change, and the type of vehicle.

[0098] The central controller can first determine whether the road resources in front of the current driving lane are occupied and whether the geometry of the road in front has changed according to the position of the "current driving lane". If at least one of the judgment results is yes, it determines that the area in front of the target vehicle is a critical conflict area and triggers the forced lane change logic, such as vehicle B in Figure 2(a) and vehicle G in Figure 2(b). The lane change logic can also be triggered according to the "active lane change request" in the current driving environment parameters, such as vehicle E in Figure 2(a).

[0099] The central controller may receive current driving environment parameters from multiple vehicles at the same time. However, when multiple vehicles need to change lanes and the front is not a critical conflict zone, the vehicle that is in the front should be considered and the lane change logic should be triggered first. As shown in Figure 2(a), assuming that the front of vehicles A to E is not a critical conflict zone, since vehicle A is in the front of these vehicles, the lane change logic is triggered first for vehicle A to allocate road rights. However, if only vehicles E and D send the current driving environment parameters, since vehicle D is in front of vehicle E, the lane change logic is triggered first for vehicle D.

[0100] For step S102, this solution also optimizes the trajectories of all vehicles based on the traffic order. To eliminate infeasible traffic orders, the following assumptions are made: 1) Due to road geometry constraints or traffic rules, vehicles that must change lanes have only one possible action (such as changing lanes), and vehicles that have a negative impact on lane change have only one possible action (such as going straight), and the remaining vehicles can choose to go straight or change lanes at will; 2) For other vehicles in the same lane, the front vehicle should have a higher traffic priority than all the following vehicles. In addition to the above basic assumptions, other reasonable assumptions can be customized according to the characteristics of the actual application scenario, and this solution does not limit them here.

[0101] A vehicle may have two possible actions, so taking the target vehicle as the starting point, multiple traffic orders are generated by randomly selecting vehicles and the possible actions of each vehicle among other vehicles around the target vehicle. The traffic order represents the priority of each vehicle, and vehicles with higher rankings have the right to occupy road resources first. The traffic order is usually expressed in the form of a string, such as CAB, which means that vehicle C has the highest priority of right of way. When other vehicles AB conflict with it, they need to slow down and give way. When there is no conflict between vehicles A and C, the traffic orders CAB and ACB have the same performance.

[0102] Taking Figure 2(a) as an example, vehicle B needs to execute the lane change logic. According to its current driving environment parameters, it is determined that the front vehicle in the right lane change target lane is vehicle A and the rear vehicle is vehicle C. There are no front and rear vehicles in the current lane of vehicle B. Therefore, a local conflict zone can be constructed by the areas where vehicles A, B and C are located. Or the lane change of vehicle B only has a negative impact on vehicle C, and the local conflict zone can be constructed only by the areas where vehicles B and C are located. Only the feasible actions of each vehicle in the local conflict zone are considered, such as vehicle A going straight or changing lanes, vehicle B only changing lanes, and vehicle C only going straight, to generate multiple traffic sequences ABC, ACB, BAC, BCA, CAB, CBA, AB, AC, BA, BC, CA, CB. Since the front of vehicle B is a key conflict zone, the right of way is allocated to vehicle B first, and four traffic sequences of BAC, BCA, BA, and BC are selected.

[0103] For step S103, most complex traffic scenes are composed of basic scene elements, including intersections, on- and off-ramps, roundabouts, and road sections. As shown in Figures 2(a) and 2(b), most basic traffic scenes contain a downstream key conflict zone and multiple local conflict zones. Therefore, an urban road network can be divided into two typical traffic scenes, intersection scenes and road section scenes, according to the geometric characteristics of the roads. At this time, the movement of the vehicle can be regarded as driving from one traffic scene to another, and the lane change control in each traffic scene can be operated separately.

[0104] Therefore, if all typical traffic scenarios can be handled well, the control and coordination of more complex traffic scenarios can be achieved through expansion and combination. This scheme proposes a group decision-making based on a two-level planning framework. The upper level of this method aims to find the optimal target passage sequence to maximize traffic efficiency, while the lower level aims to resolve the driving trajectory conflict of each vehicle in the local conflict area according to the target passage sequence.

[0105] The problem of finding the optimal target passage sequence is essentially a tree search problem. In order to speed up the search process, the upper layer of the group decision-making based on the two-level planning framework preferably adopts a combination of the Monte Carlo tree search method and heuristic rules. Through the heuristic rules, one or more passage rules that meet the first-in-first-out principle are screened out from multiple passage rules, and it is ensured that the obtained passage sequence is basically reasonable. Then, the Monte Carlo tree search method is used to continue screening from these one or more passage rules.

[0106] Compared with the existing method of traversing all feasible pass sequences using the optimal planning strategy, the Monte Carlo tree search method is more inclined to search for a pass sequence that has the potential to become the optimal one (i.e., the target pass sequence), thereby greatly improving the overall search efficiency. During the construction of the search tree, the current optimal target pass sequence is continuously updated. When the maximum computing time budget is reached, the search process ends and returns the currently selected target pass sequence.

[0107] In addition, since structured roads restrict vehicle movement, sometimes the right of way allocation can only be adjusted within a limited space. By using heuristic rules to prioritize the order of traffic, and then continuously adjusting the right of way of some vehicles in the order, a better solution can be obtained. These properties also ensure that the decision method can find at least one feasible solution for any scenario.

[0108] Based on the lower layer of the group decision-making of the two-level planning framework, an interpretation algorithm from traffic order to trajectory is designed. This interpretation algorithm can quickly and accurately plan conflict-free trajectories for each vehicle according to the vehicle priority order in the target traffic order. Vehicles with higher priorities will be given priority in planning trajectories. The traffic order given by the upper layer planning is evaluated accordingly. If the lower layer cannot find the corresponding lane change trajectory for the target traffic order, a value will be fed back to the upper layer to indicate that the traffic order is not feasible. After receiving the feedback value, the upper layer will adjust the search direction by the Monte Carlo tree search method to avoid generating similar traffic orders again.

[0109] By iteratively solving the upper and lower layer problems, the current optimal traffic order is continuously updated. In actual operation, the interpretation algorithm is relatively complex to run, so it is difficult to find the optimal solution within a certain period of time. Considering that suboptimal solutions usually do not significantly reduce the effect, and in order to ensure the real-time performance of the overall solution process, it is preferred to obtain the optimal approximate solution in a shorter time rather than spending a lot of time to obtain the global optimal solution. When the maximum allowed computing time limit is reached, the search process ends and returns the current optimal target traffic order and the vehicle conflict-free trajectory corresponding to the current optimal target traffic order.

[0110] For step S104, it is difficult to consider multiple feasible lane-changing trajectories in the existing optimization problem, so it is generally assumed that the lane-changing trajectories of all vehicles are the same. However, there is only one feasible lane-changing trajectory, which often leads to an unsolvable problem. To solve this problem, this solution pre-constructs a trajectory set, which includes a large number of different lane-changing trajectories, and each trajectory is generated according to constraints such as vehicle type, initial velocity, acceleration, and final velocity.

[0111] Each vehicle can search for multiple matching conflict-free trajectories from the trajectory set based on its own driving parameters, such as the above-mentioned vehicle type, initial velocity, acceleration and other constraints, as shown in Figure 2(c), and select a conflict-free trajectory with the highest right of way priority, so that the target vehicle and other vehicles around the target do not collide when driving at the same time.

[0112] For the target vehicle, after selecting a conflict-free trajectory with the highest right of way priority, the road time-space occupancy information is updated according to the conflict-free lane-changing trajectory of the target vehicle, so that other subsequent vehicles can make lane-changing decisions based on the updated road time-space occupancy information.

[0113] The method provided in the above embodiment adopts a combination of Monte Carlo tree search method and heuristic rules to speed up the efficiency of screening the target traffic sequence and reduce the amount of calculation; the corresponding trajectory can be matched from the preset trajectory set according to the vehicle driving parameters, thereby improving the efficiency of lane change decision generation, which is of great significance for ensuring road traffic safety and improving road traffic capacity.

[0114] See also Figure 3 , shows a schematic flow chart of an optional vehicle lane change processing method according to an embodiment of the present invention, comprising the following steps:

[0115] S301: calling a Monte Carlo tree search program to determine a node with an unexpanded action in each selected passage sequence, and creating a new node as a child node of the node; wherein the child node is used to indicate a driving environment parameter of the target vehicle after executing the unexpanded action;

[0116] S302: starting from the child node, sampling from vehicles not included in each of the passing sequences, adding the sampled vehicles to each of the passing sequences, repeating the above random sampling operation until a leaf node is generated, and obtaining a simulation score;

[0117] S303: The node in each passage sequence and all nodes on the path from the root node to the node are scored cumulatively according to the simulation score; wherein the root node is used to indicate the current driving environment parameters of the target vehicle;

[0118] S304: Calculate the confidence value according to the number of simulations and the accumulated score of each node, determine the target node with the largest confidence value, and use the transit sequence between the target node and the root node as the target transit sequence.

[0119] In the above implementation, for steps S301 to S303, taking FIG2(a) as an example, the road segment scene includes vehicles A, B, C, D and E, thus forming a set {A, B, C, D, E}, and the total number of permutations of the set is 5! = 120. Therefore, when there are a large number of vehicles in the local conflict zone that need to consider lane changing, the number of traffic orders to be calculated will be huge. At this time, it is impossible to expand all leaf nodes in a search tree within a limited computing time. Therefore, this solution preferably uses the Monte Carlo tree search method MCTS to search for the optimal target traffic order from the multiple traffic orders generated above.

[0120] The MCTS method constructs a search tree in an iterative manner. Each iteration usually includes four steps: selection, expansion, simulation, and back propagation. Each node in the search tree contains three basic information: driving environment parameters, simulation times, and cumulative scores.

[0121] 1. Selection stage: Each expanded node may have the following possibilities:

[0122] 1) If all the possible actions of the node have been expanded, calculate the UCT values ​​of all nodes of the node, find the child node with the largest UCT value, and continue to check, repeatedly iterating downward.

[0123] 2) The node has available actions that have not been expanded. For example, if a node has 20 available actions, but only 19 child nodes are created in the search tree, then this node is the node of this iteration, and the actions that have not been expanded are determined.

[0124] 2. Expansion phase: At the end of the selection phase, find the node Y that is most urgently expanded and an action X that has not yet been expanded. Create a new node Yn in the search tree as a new child node of Y. The situation of Yn is the situation of node Y after executing action X, and for this solution, it is the driving environment parameters of the target vehicle.

[0125] 3. Simulation stage: In order to give Yn an initial score, randomly sample vehicles that are not included in the passage order starting from Yn until an outcome is obtained. This outcome will be used as Yn's initial score (i.e. simulation score). Generally, victory or failure is used as the score, which can only be 1 or 0.

[0126] 4. Back propagation phase: After the simulation of Yn is completed, its parent node Y and all nodes on the path from the root node to node Y will add their own cumulative scores based on the results of this simulation. Each iteration will expand the search tree, and as the number of iterations increases, the size of the search tree will continue to increase. When a certain number of iterations or time is reached, the best child node under the root node is selected as the result of this decision.

[0127] For step S304, the MCTS method continuously simulates each move until the end, and the winning rate of the move is calculated as W / N based on the total number of simulations N and the number of successful simulation results W. In order to maintain a balance between the maximum winning rate and the exploration of new nodes, UCT (Upper Confidence Bound) is introduced.

[0128] The so-called confidence interval is the credibility of the probability calculation result. UCT can be used to correct the problem of too few samples. For example, if a coin is flipped three times and all of them are heads up, the probability of the coin flipping heads up is calculated to be 100%. The reason is that the sample is too small. The formula used by UCT is:

[0129]

[0130] Among them, w i is the number of times the simulation result of node i is successful, that is, the cumulative score, n i is the number of simulations for node i, N i is the number of simulations for all nodes, c is the exploration constant, and the theoretical value is Can be adjusted based on experience.

[0131] Therefore, after the iteration is completed, the node Y on each pass order and all nodes on the path from the root node to node Y input the two values ​​of the cumulative score and the number of simulations obtained in the last iteration into the above formula to obtain the UCT value of each node. The node with the largest UCT value is taken as the target node, and the pass order between the target node and the root node is taken as the optimal target pass order.

[0132] See also Figure 4 As shown, at the beginning, the search tree has only one root node R, which represents the current driving environment parameters of the target vehicle, and the root node R is used as the starting node of each traffic sequence. See Figure 2 (a) Assuming that the current local conflict area includes vehicles A, B, and C, according to the random selection of vehicles and the available actions of each vehicle, and the front of vehicle B is the key conflict area, the right of way is allocated to vehicle B first, and four traffic sequences of BAC, BCA, BA, and BC are selected. Since vehicles A and C are in the same lane, two traffic sequences are finally obtained: BA and BC.

[0133] In addition to changing lanes to vehicle A, vehicle B can also change lanes to vehicle C. Assuming that there is only one traffic sequence BA, node BC can be expanded. Continuing with node BC as the starting point, a direct child node is randomly generated at its lower layer using the random sampling method, assuming it is BCA. If there are other vehicles, node BCA will repeat the above process until a leaf node is generated, such as node BCADE. After the leaf node is generated, a simulation result will be obtained, that is, the simulation is successful or the simulation fails.

[0134] The classic MCTS method randomly samples vehicles that are not yet included in the traffic order, and then continuously adds the sampled vehicles to the existing traffic order until a complete traffic order is formed. At the same time, no branches are generated in this process, but it goes down until the maximum depth of the tree is reached. During the construction of the search tree, the current optimal traffic order is continuously updated, and the search process ends when the upper limit of the number of iterations is reached or the maximum computing time budget is reached (that is, the iteration time reaches the preset time). Based on the optimal target traffic order, the speed and acceleration curves of all vehicles can be calculated through the traffic order to trajectory interpretation algorithm in the lower-level planning.

[0135] The method provided in the above embodiment uses the Monte Carlo tree search method to convert the search problem of determining the optimal target traffic order into a tree search problem of constructing all feasible traffic orders. By simulating multiple vehicles and their feasible actions, the node with the highest confidence value is selected to improve the accuracy of determining the target traffic order.

[0136] See also Figure 5, shows a schematic flow chart of another optional vehicle lane change processing method according to an embodiment of the present invention, comprising the following steps:

[0137] S501: Calculating a safe distance between the target vehicle and the vehicle in front of the lane-changing target lane so as not to collide with each other according to the driving parameters of the target vehicle and the driving parameters of the vehicle in front of the lane-changing target lane;

[0138] S502: In response to the distance between the target vehicle and the preceding vehicle in the lane-changing target lane being less than the safe vehicle distance, adjusting the speed of the target vehicle so that the distance between the target vehicle and the preceding vehicle in the lane-changing target lane is greater than or equal to the safe vehicle distance;

[0139] S503: Repeat the above operation of adjusting the speed to meet the collision constraint condition, so that the distance between the target vehicle and each other vehicle and the corresponding preceding vehicle all meets the collision constraint condition.

[0140] In the above implementation, for steps S501 to S503, after the target traffic sequence is screened out based on the Monte Carlo tree search method, a conflict-free trajectory can be planned for each vehicle. Therefore, the decision variables at this time become: target traffic sequence, expected arrival time, speed and acceleration curve of each vehicle, and vehicle distance. Speed, acceleration, and vehicle distance can be located in the vehicle's driving parameters.

[0141] This scheme designs an algorithm for interpreting the passage sequence to the trajectory, see Figure 6 As shown in the figure, the target vehicle adjusts the distance with the vehicle in front of the lane change target lane by accelerating and decelerating to ensure that there is sufficient space to change lanes. This is a distance-based method to judge the collision risk. In addition, there are many other types of methods that can be used to evaluate the collision risk during lane change, such as the Time To Collision (TTC)-based method and the Risk Index-Based method, which can be combined into this solution. This solution can also use collision avoidance constraints, which are defined as follows:

[0142] d ij (t)≥F ij (t)

[0143]

[0144]

[0145] Among them, v i (t) is the speed of vehicle i, v j (t) is the speed of vehicle j, d ij (t) is the distance between vehicle i and vehicle j, F ij(t) is the safe headway, which indicates the length of the prohibited driving area between vehicle i and vehicle j. The prohibited driving area refers to the time and space area where the front vehicle has absolute right of way, and the rear vehicle cannot enter the prohibited driving area owned by the front vehicle; ρ is the safe headway, a i,brake (t) is the average braking acceleration of vehicle i, a i,max,brake is the maximum braking acceleration of vehicle i, a i,min,brake is the minimum braking acceleration of vehicle i, where the subscript m represents the maximum or minimum, which can be selected arbitrarily. In addition, since parameter a i,max,brake 、a i,min,brake It is related to the type of vehicle, so the safe distance F ij (t) It also depends on the vehicle type, that is, this solution is applicable to scenarios consisting of different types of vehicles.

[0146] It should be noted that the preceding vehicle here includes the preceding vehicle in the current lane of the vehicle and the preceding vehicle in the lane of the lane change target, such as Figure 6 In the process, the longitudinal speed of vehicles that do not meet the collision avoidance constraint conditions is adjusted until the safe lane change conditions are met. Vehicle I is changing lanes to the right lane, so its front vehicle is vehicle J. The distance between vehicle I and vehicle J must meet d ij (t)≥F ij (t). After changing lanes, vehicle I is located in front of vehicle K, so the vehicle in front of vehicle K is vehicle I, and the distance between vehicle K and vehicle I must satisfy d ki (t)≥F ki (t). In the lane change lane, vehicle K is ahead of vehicle J, so the distance between the two vehicles must satisfy d kj (t)≥F kj (t), but due to d ij (t)≥F ij (t), d ki (t)≥F ki (t), the distance between vehicle K and vehicle J must satisfy the collision constraint condition, so it is preferred not to consider the distance between vehicle K and vehicle J to reduce the amount of calculation. Similarly, for the vehicle in the lane where vehicle I is located and located in front of vehicle I, the distance between vehicle I and vehicle I is not considered. The default distance satisfies the collision constraint condition.

[0147] All unplanned vehicles follow the vehicle in front according to the following algorithm, such as Newell following algorithm, following algorithm based on mass-spring-damper-clutch model, etc., to control the unplanned vehicles to maintain a safe and appropriate distance from the vehicle in front. The following algorithm usually acts on the first vehicle in the front position on the road section, such as Figure 6 Vehicle J.

[0148] In actual vehicle driving, the relative distance between the target vehicle and the vehicle in front of the current lane, the target vehicle and the vehicle in front of the lane change target lane, and the target vehicle and the vehicle behind the lane change target lane has a greater impact on the lane change behavior decision than the influence of the relative speed of the two vehicles. If the relative distance between the target vehicle and the vehicle in front of the current lane and the vehicle in front of the lane change target lane is too large, the target vehicle needs to adopt an acceleration strategy to adjust the speed to avoid excessive vehicle distance. If the relative distance between the target vehicle and the vehicle behind the lane change target lane is too small, the target vehicle needs to adopt a deceleration strategy to adjust the vehicle distance. When the vehicle distances between the target vehicle and the vehicle in front of the current lane, the target vehicle and the vehicle in front of the lane change target lane, and the target vehicle and the vehicle behind the lane change target lane all meet the collision constraint conditions, the target vehicle adopts a uniform speed strategy.

[0149] The method provided in the above embodiment can effectively identify situations where the collision constraint conditions are not satisfied between the vehicle and the surrounding environment, and construct a safe lane changing scenario by adjusting the distance between vehicles through acceleration and deceleration operations, thereby ensuring the safety of the lane changing process, and has strong application significance in actual road driving.

[0150] This solution aims at the decision-making problem of road scene groups considering lane change. The proposed solution analyzes the necessity and safety of autonomous or forced lane change of autonomous driving vehicles, effectively reducing the solution complexity brought by lane change trajectory:

[0151] 1. Determine whether the road ahead of the vehicle is a critical conflict zone based on whether the road resources ahead of the vehicle are occupied or whether the road geometry has changed, and determine whether to execute the forced lane change logic based on the judgment result; or execute the lane change logic based on the lane change request transmitted by the vehicle;

[0152] 2. A group decision-making approach based on two-level planning is proposed. The upper level of this approach quickly selects the target traffic sequence from multiple traffic sequences by combining the Monte Carlo tree search method with heuristic rules, thereby improving the solution efficiency. The lower level of this approach focuses on dealing with local conflicts caused by behaviors such as lane changing, designs an algorithm for interpreting the traffic sequence to the trajectory, adjusts the speed of each vehicle so that the vehicle distance meets the collision constraint conditions, and facilitates the rapid planning of conflict-free trajectories for all vehicles, thereby achieving efficient coordination of the vehicle group.

[0153] See also Figure 7 , which shows a schematic diagram of main modules of a vehicle lane change processing device 700 provided in an embodiment of the present invention, including:

[0154] The determination module 701 is used to receive the current driving environment parameters transmitted by the target vehicle to determine the information of other vehicles located around the target vehicle; optionally, it can also be used to determine whether a lane change operation needs to be performed. For example, if the road resources in front of the current driving lane of the target vehicle in the current driving environment parameters are occupied or the road geometry changes, the lane change logic is triggered; or if the current driving environment parameters include a lane change request transmitted by the target vehicle, the lane change logic is triggered.

[0155] The generation module 702 is used to generate multiple passage sequences by selecting vehicles and corresponding vehicle information from the other vehicles with the target vehicle as the starting point; specifically, multiple passage sequences are generated by randomly selecting vehicles and each vehicle's movable action.

[0156] The processing module 703 is used to select a target passing order from the multiple passing orders, and plan a conflict-free trajectory for each vehicle in turn according to the vehicle priority order in the target passing order, so that the target vehicle and the other vehicles do not collide when traveling at the same time; specifically:

[0157] 1) The implementation steps for selecting a target passing sequence from the plurality of passing sequences are as follows: calling a heuristic rule, and selecting one or more passing sequences from the plurality of passing sequences according to the first-in-first-out principle based on the current positions of the target vehicle and each other vehicle;

[0158] Calling a Monte Carlo tree search program to determine a node with an unexpanded action in each selected passage sequence, and creating a new node as a child node of the node; wherein the child node is used to indicate a driving environment parameter of the target vehicle after executing the unexpanded action;

[0159] Taking the child node as the starting point, sampling is performed from vehicles that are not included in each of the passing sequences, the sampled vehicles are added to each of the passing sequences, and the above random sampling operation is repeated until a leaf node is generated, thereby obtaining a simulation score;

[0160] The nodes in each passage sequence and all nodes on the path from the root node to the nodes are scored cumulatively according to the simulation score; wherein the root node is used to indicate the current driving environment parameters of the target vehicle;

[0161] The confidence value is calculated according to the number of simulations and the cumulative score of each node, the target node with the largest confidence value is determined, and the passage order between the target node and the root node is used as the target passage order.

[0162] 2) For planning a conflict-free trajectory, the following steps are performed: based on the driving parameters of the target vehicle and the driving parameters of the vehicle in front of the target lane, a safe distance between the target vehicle and the vehicle in front of the target lane is calculated so as to avoid collision;

[0163] In response to a distance between the target vehicle and a preceding vehicle in the lane-changing target lane being less than a safe vehicle distance, adjusting a speed of the target vehicle so that the distance between the target vehicle and a preceding vehicle in the lane-changing target lane is greater than or equal to the safe vehicle distance;

[0164] The above operation of adjusting the speed to meet the collision constraint condition is repeated so that the distance between the target vehicle and each other vehicle and the corresponding preceding vehicle all meets the collision constraint condition.

[0165] According to the driving parameters of each vehicle, searching for a conflict-free trajectory matching the driving parameters from a preset trajectory set; wherein the driving parameters include speed, acceleration, and vehicle type;

[0166] When there are multiple matched conflict-free trajectories, the conflict-free trajectory with the highest right-of-way priority among the multiple conflict-free trajectories is used as the target conflict-free trajectory.

[0167] The lane changing module 704 is used to trigger the target vehicle to perform a lane changing operation according to the conflict-free trajectory planned for the target vehicle.

[0168] The implementation device of the present invention further includes a lane change priority determination module, which is used to:

[0169] When receiving current driving environment parameters transmitted by multiple vehicles, priority is given to allocating road rights to vehicles that are in a critical conflict zone ahead; wherein the critical conflict zone is an area where road resources are occupied or road geometry has changed; and

[0170] When there is no critical conflict zone ahead of each vehicle, the right of way is allocated preferentially to the vehicle at the front among the multiple vehicles.

[0171] In the implementation device of the present invention, the determination module is further used for:

[0172] Based on the current areas where the target vehicle and the other vehicles are located, a local conflict area is constructed, and an available action of each other vehicle in the local conflict area is determined.

[0173] In addition, the specific implementation content of the device described in the embodiment of the present invention has been described in detail in the method described above, so the repeated content will not be described again here.

[0174] Figure 8An exemplary system architecture 800 to which embodiments of the present invention may be applied is shown, including terminal devices 801 , 802 , 803 , a network 804 and a server 805 (only an example).

[0175] Terminal devices 801, 802, 803 may be various electronic devices with display screens and supporting web browsing, and may be installed with various communication client applications. Users may use terminal devices 801, 802, 803 to interact with server 805 via network 804 to receive or send messages, etc.

[0176] The network 804 is used to provide a medium for communication links between the terminal devices 801, 802, 803 and the server 805. The network 804 may include various connection types, such as wired, wireless communication links or optical fiber cables.

[0177] The server 805 may be a server that provides various services. It should be noted that the method provided in the embodiment of the present invention is generally executed by the server 805 , and accordingly, the device is generally set in the server 805 .

[0178] It should be understood that Figure 8 The number of terminal devices, networks and servers in the embodiment is only for illustration. Any number of terminal devices, networks and servers may be provided according to the implementation requirements.

[0179] Reference below Fig. 9 , which shows a schematic diagram of the structure of a computer system 900 of a terminal device suitable for implementing an embodiment of the present invention. Fig. 9 The terminal device shown is only an example and should not bring any limitation to the functions and scope of use of the embodiments of the present invention.

[0180] like Fig. 9 As shown, the computer system 900 includes a central processing unit (CPU) 901, which can perform various appropriate actions and processes according to a program stored in a read-only memory (ROM) 902 or a program loaded from a storage part 908 into a random access memory (RAM) 903. In the RAM 903, various programs and data required for the operation of the system 900 are also stored. The CPU 901, the ROM 902, and the RAM 903 are connected to each other via a bus 904. An input / output (I / O) interface 905 is also connected to the bus 904.

[0181] The following components are connected to the I / O interface 905: an input section 906 including a keyboard, a mouse, etc.; an output section 907 including a cathode ray tube (CRT), a liquid crystal display (LCD), etc., and a speaker, etc.; a storage section 908 including a hard disk, etc.; and a communication section 909 including a network interface card such as a LAN card, a modem, etc. The communication section 909 performs communication processing via a network such as the Internet. A drive 910 is also connected to the I / O interface 905 as needed. A removable medium 911, such as a magnetic disk, an optical disk, a magneto-optical disk, a semiconductor memory, etc., is installed on the drive 910 as needed, so that a computer program read therefrom is installed into the storage section 908 as needed.

[0182] In particular, according to the embodiments disclosed in the present invention, the process described above with reference to the flowchart can be implemented as a computer software program. For example, the embodiments disclosed in the present invention include a computer program product, which includes a computer program carried on a computer-readable medium, and the computer program includes a program code for executing the method shown in the flowchart. In such an embodiment, the computer program can be downloaded and installed from the network through the communication part 909, and / or installed from the removable medium 911. When the computer program is executed by the central processing unit (CPU) 901, the above-mentioned functions defined in the system of the present invention are executed.

[0183] It should be noted that the computer-readable medium shown in the present invention may be a computer-readable signal medium or a computer-readable storage medium or any combination of the above two. The computer-readable storage medium may be, for example, but not limited to, an electrical, magnetic, optical, electromagnetic, infrared, or semiconductor system, device or device, or any combination of the above. More specific examples of computer-readable storage media may include, but are not limited to: an electrical connection with one or more wires, a portable computer disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the above. In the present invention, a computer-readable storage medium may be any tangible medium containing or storing a program that can be used by or in combination with an instruction execution system, device or device. In the present invention, a computer-readable signal medium may include a data signal propagated in a baseband or as part of a carrier wave, which carries a computer-readable program code. This propagated data signal may take a variety of forms, including but not limited to electromagnetic signals, optical signals, or any suitable combination of the above. Computer-readable signal media may also be any computer-readable medium other than computer-readable storage media, which may send, propagate or transmit a program for use by or in conjunction with an instruction execution system, apparatus or device. The program code contained on the computer-readable medium may be transmitted using any appropriate medium, including but not limited to: wireless, wire, optical cable, RF, etc., or any suitable combination of the above.

[0184] The flow chart and block diagram in the accompanying drawings illustrate the possible architecture, function and operation of the system, method and computer program product according to various embodiments of the present invention. In this regard, each box in the flow chart or block diagram can represent a module, a program segment, or a part of a code, and the above-mentioned module, program segment, or a part of a code contains one or more executable instructions for realizing the specified logical function. It should also be noted that in some alternative implementations, the functions marked in the box can also occur in a different order from the order marked in the accompanying drawings. For example, two boxes represented in succession can actually be executed substantially in parallel, and they can sometimes be executed in the opposite order, depending on the functions involved. It should also be noted that each box in the block diagram or flow chart, and the combination of the boxes in the block diagram or flow chart can be implemented with a dedicated hardware-based system that performs a specified function or operation, or can be implemented with a combination of dedicated hardware and computer instructions.

[0185] The modules involved in the embodiments of the present invention may be implemented by software or hardware. The modules described may also be set in a processor. For example, they may be described as follows: a processor includes a determination module, a generation module, a processing module, and a lane change module. The names of these modules do not, in some cases, constitute limitations on the modules themselves. For example, the generation module may also be described as a "traffic sequence generation module."

[0186] As another aspect, the present invention further provides a computer-readable medium, which may be included in the device described in the above embodiment; or may exist independently without being assembled into the device. The above computer-readable medium carries one or more programs, and when the above one or more programs are executed by a device, the device executes any of the above vehicle lane change processing methods.

[0187] The computer program product of the present invention comprises a computer program, and when the computer program is executed by a processor, the vehicle lane changing processing method in the embodiment of the present invention is implemented.

[0188] The above specific implementations do not constitute a limitation on the protection scope of the present invention. It should be understood by those skilled in the art that various modifications, combinations, sub-combinations and substitutions may occur depending on design requirements and other factors. Any modification, equivalent substitution and improvement made within the spirit and principle of the present invention shall be included in the protection scope of the present invention.

Claims

1. A vehicle lane change processing method, characterized in that: include: Receiving current driving environment parameters transmitted by the target vehicle to determine information of other vehicles located around the target vehicle; Taking the target vehicle as a starting point, a plurality of passing orders are generated by selecting vehicles and corresponding vehicle information from the other vehicles; wherein the plurality of passing orders are generated by randomly selecting vehicles and the movable actions of each vehicle; A target traffic sequence is selected from the multiple traffic sequences, and a conflict-free trajectory is planned for each vehicle in turn according to the vehicle priority order in the target traffic sequence, so that the target vehicle and the other vehicles do not collide when traveling at the same time; wherein the group decision based on the two-layer planning framework includes an upper layer and a lower layer, the upper layer selects the target traffic sequence by combining the Monte Carlo tree search method and the heuristic rule, and the lower layer is used to plan the conflict-free trajectory and evaluate the target traffic sequence selected by the upper layer; the upper and lower layer problems are iteratively solved to update the current optimal traffic sequence; According to the conflict-free trajectory planned for the target vehicle, the target vehicle is triggered to perform a lane change operation.

2. The method according to claim 1, characterized in that Before determining the information of other vehicles located around the target vehicle, the method further includes: If, in the current driving environment parameters, the road resources ahead of the lane where the target vehicle is currently traveling are occupied or the road geometry changes, then the lane change logic is triggered; or If the current driving environment parameters include a lane change request transmitted by the target vehicle, the lane change logic is triggered.

3. The method according to claim 2, characterized in that Also includes: When receiving current driving environment parameters transmitted by multiple vehicles, the right of way is allocated to vehicles in the critical conflict zone ahead. The critical conflict zone is an area where road resources are occupied or the road geometry changes. as well as When there is no critical conflict zone ahead of each vehicle, the right of way is allocated preferentially to the vehicle at the front among the multiple vehicles.

4. The method according to any one of claims 1 to 3, characterized in that The determining of information of other vehicles located around the target vehicle also includes: Based on the current areas where the target vehicle and the other vehicles are located, a local conflict area is constructed, and an available action of each other vehicle in the local conflict area is determined.

5. The method according to claim 1, characterized in that The step of selecting a target pass sequence from the plurality of pass sequences includes: Invoke a heuristic rule to select one or more passing sequences from the plurality of passing sequences according to a first-in-first-out principle based on the current positions of the target vehicle and each other vehicle; Calling a Monte Carlo tree search program to determine a node having an unexpanded action in each selected passage sequence, and creating a new node as a child node of the node; wherein the child node is used to indicate a driving environment parameter of the target vehicle after executing the unexpanded action; Taking the child node as the starting point, sampling is performed from vehicles that are not included in each of the passage sequences, the sampled vehicles are added to each of the passage sequences, and the random sampling operation is repeated until a leaf node is generated, thereby obtaining a simulation score; The nodes in each passage sequence and all nodes on the path from the root node to the nodes are scored cumulatively according to the simulation score; wherein the root node is used to indicate the current driving environment parameters of the target vehicle; The confidence value is calculated according to the number of simulations and the cumulative score of each node, the target node with the largest confidence value is determined, and the passage order between the target node and the root node is used as the target passage order.

6. The method according to claim 1, characterized in that Before planning a conflict-free trajectory for each vehicle in turn according to the vehicle priority order in the target traffic order, the method further includes: Calculating a safe distance between the target vehicle and the vehicle in front of the lane-changing target lane so that the target vehicle does not collide with the vehicle in front of the lane-changing target lane according to the driving parameters of the target vehicle and the driving parameters of the vehicle in front of the lane-changing target lane; In response to a distance between the target vehicle and a preceding vehicle in the lane-changing target lane being less than a safe vehicle distance, adjusting a speed of the target vehicle so that the distance between the target vehicle and a preceding vehicle in the lane-changing target lane is greater than or equal to the safe vehicle distance; The speed is adjusted repeatedly to satisfy the collision constraint condition, so that the distance between the target vehicle and each other vehicle and the corresponding preceding vehicle satisfies the collision constraint condition.

7. The method according to claim 1 or 6, characterized in that: The step of planning a conflict-free trajectory for each vehicle in turn includes: According to the driving parameters of each vehicle, searching for a conflict-free trajectory matching the driving parameters from a preset trajectory set; wherein the driving parameters include speed, acceleration, and vehicle type; When there are multiple matched conflict-free trajectories, the conflict-free trajectory with the highest right of way priority among the multiple conflict-free trajectories is used as the target conflict-free trajectory.

8. A vehicle lane change processing device, characterized in that: include: A determination module, used for receiving current driving environment parameters transmitted by a target vehicle to determine information of other vehicles located around the target vehicle; A generation module is used to generate multiple passage sequences by selecting vehicles and corresponding vehicle information from the other vehicles, taking the target vehicle as a starting point; wherein the multiple passage sequences are generated by randomly selecting vehicles and each vehicle's movable action; A processing module is used to select a target traffic sequence from the multiple traffic sequences, and plan a conflict-free trajectory for each vehicle in turn according to the vehicle priority order in the target traffic sequence, so that the target vehicle and the other vehicles do not collide when traveling at the same time; wherein the group decision based on the two-layer planning framework includes an upper layer and a lower layer, the upper layer uses a Monte Carlo tree search method and a heuristic rule to select the target traffic sequence, and the lower layer is used to plan a conflict-free trajectory and evaluate the target traffic sequence selected by the upper layer; by iteratively solving the upper and lower layer problems, the current optimal traffic sequence is updated; The lane changing module is used to trigger the target vehicle to perform a lane changing operation according to the conflict-free trajectory planned for the target vehicle.

9. An electronic device, characterized in that: include: one or more processors; a storage device for storing one or more programs, When the one or more programs are executed by the one or more processors, the one or more processors implement the method according to any one of claims 1 to 7.

10. A computer readable medium having a computer program stored thereon, characterized in that: When the program is executed by a processor, the method according to any one of claims 1 to 7 is implemented.

Images