Vehicle trajectory planning method and electronic device

Abstract

This application discloses a vehicle trajectory planning method and electronic device. The method includes: determining multiple first predicted trajectories of target obstacles surrounding a target vehicle based on first target information; determining multiple second predicted trajectories corresponding to the target vehicle based on second target information including the current driving state information of the target vehicle; and determining multiple third predicted trajectories corresponding to the target vehicle based on map information, the second target information, and third target information including the current operating state information of the target obstacles. A first evaluation result is determined for each first predicted trajectories, and a second evaluation result is determined for each second and third predicted trajectories based on the first evaluation result and evaluation indicators. Finally, a target trajectory corresponding to the target vehicle is determined based on the second evaluation result and trajectory planning continuity requirements. This effectively improves the accuracy of the target vehicle's target trajectory.

Description

Technical Field

[0001] This application relates to the field of vehicle trajectory planning technology, and in particular to a vehicle trajectory planning method and electronic device. Background Technology

[0002] With the development of vehicle intelligence, assisted driving performance has become an important indicator for vehicle performance evaluation. During assisted driving, the vehicle's trajectory needs to be planned based on the surrounding environment, and the vehicle is typically guided to execute the optimal trajectory.

[0003] Existing trajectory planning methods typically rely on rule-based candidate trajectory prediction algorithms, assuming all other vehicles are moving at constant speeds. However, in reality, vehicles do not move at constant speeds in most traffic scenarios, and the scenarios themselves are often complex. Because these methods rely on overly simplistic assumptions of constant speed and rule-based candidate trajectory prediction, they are poorly suited for complex driving environments, particularly in dynamic, interactive, and non-linear driving conditions. The accuracy of the determined trajectories is inconsistent, impacting vehicle efficiency, safety, and user experience. Summary of the Invention

[0004] This application provides a vehicle trajectory planning method to address the problems in existing assisted driving trajectory planning methods, which are typically rule-based candidate trajectory prediction algorithms. These methods obtain the vehicle's trajectory under the pre-set condition that other vehicles are moving at a constant speed. The accuracy of the vehicle trajectory determined by this method is unstable, affecting vehicle driving efficiency, vehicle driving safety, and user experience.

[0005] To address the aforementioned technical problems, in a first aspect, this application discloses a vehicle trajectory planning method applied to a trajectory planning model. The trajectory planning model includes an obstacle trajectory prediction module, a vehicle trajectory prediction module, a trajectory evaluation module, and a trajectory determination module. The vehicle trajectory prediction module includes a rule-based trajectory planner and a learning-based trajectory planner. The method includes: the obstacle trajectory prediction module determining multiple first predicted trajectories of target obstacles surrounding the target vehicle based on first target information; the vehicle trajectory prediction module determining multiple second predicted trajectories corresponding to the target vehicle based on second target information using a rule-based trajectory planner; and determining multiple third predicted trajectories corresponding to the target vehicle based on third target information using a learning-based trajectory planner; the trajectory evaluation module determining a first evaluation result corresponding to each first predicted trajectories, and determining a second evaluation result corresponding to each second and third predicted trajectories based on the first evaluation results and evaluation indicators; and the trajectory determination module determining the target vehicle's target trajectory based on the second evaluation results and trajectory planning continuity requirements.

[0006] By adopting the above technical solution, in vehicle trajectory planning, the obstacle trajectory prediction module determines multiple first predicted trajectories of target obstacles around the target vehicle based on the first target information. Then, the vehicle trajectory prediction module determines multiple second predicted trajectories corresponding to the target vehicle based on the second target information using a rule-based trajectory planner, and multiple third predicted trajectories corresponding to the target vehicle based on the third target information using a learning-based trajectory planner. The trajectory evaluation module determines the first evaluation result corresponding to each first predicted trajectories, and determines the second evaluation results corresponding to each second and third predicted trajectories based on the first evaluation results and evaluation indicators. The trajectory determination module determines the target vehicle's target trajectory based on the second evaluation results and trajectory planning continuity requirements. Thus, in the process of vehicle trajectory planning, multiple predicted trajectories of the target vehicle can be generated from multiple aspects through rule-based and learning-based trajectory planners. Finally, based on the second evaluation results of the determined multiple predicted trajectories and the trajectory planning continuity requirements, the target vehicle's target trajectory planned from the multiple predicted trajectories is more accurate, effectively improving vehicle driving efficiency, driving safety, and user experience.

[0007] According to another specific implementation of this application, the vehicle trajectory planning method disclosed in this implementation includes: first target information including the historical trajectory information of the target vehicle, the historical trajectory information of the target obstacle, and map information of the driving environment in which the target vehicle is located; second target information including the current driving state information of the target vehicle, which includes at least one of acceleration information, speed information, distance information from the vehicle in front, desired speed information, and lane centerline information in the current driving environment; and third target information including map information of the driving environment in which the target vehicle is located, the current driving state information of the target vehicle, and the current motion state information of the target obstacle.

[0008] By adopting the above technical solution, the planned target driving trajectory of the target vehicle is more accurate by using historical running trajectory information of target obstacles around the vehicle, current running status information, multiple first predicted driving trajectories, historical running trajectory information of the target vehicle, current driving status information, and map information of the driving environment, which effectively improves vehicle driving efficiency, driving safety and user experience.

[0009] According to another specific implementation of this application, the implementation of this application discloses a vehicle driving trajectory planning method, which includes: obtaining the initial historical driving trajectory information of the target vehicle, the initial historical driving trajectory information of the target obstacle, and the initial map information of the driving environment in which the target vehicle is located.

[0010] The initial historical trajectory information of the target vehicle, the initial historical trajectory information of the target obstacle, and the initial map information are vectorized to obtain the vectorized historical trajectory information of the target vehicle, the historical trajectory information of the target obstacle, and the map information. The vectorized historical trajectory information of the target vehicle, the historical trajectory information of the target obstacle, and the map information are then filled to obtain the historical trajectory information of the target vehicle, the historical trajectory information of the target obstacle, and the map information of the driving environment of the target vehicle in tensor format.

[0011] According to another specific implementation of this application, the implementation of this application discloses a vehicle driving trajectory planning method. The vehicle trajectory prediction module determines multiple second predicted driving trajectories corresponding to the target vehicle based on second target information and through a rule-based trajectory planner. The method includes: based on the second target information, performing lateral trajectory planning processing and longitudinal trajectory planning processing through a rule-based trajectory planner to determine multiple second predicted driving trajectories corresponding to the target vehicle. The second target information includes at least one of the following: acceleration information, speed information, distance information from the vehicle in front, desired speed information, and lane centerline information in the current driving environment.

[0012] By adopting the above technical solution, based on the second target information including at least one of acceleration information, speed information, distance information from the vehicle in front, desired speed information, and lane centerline information in the current driving environment, a rule-based trajectory planner is used to perform lateral trajectory planning and longitudinal trajectory planning, which can obtain more accurate multiple second predicted driving trajectories, thereby improving the accuracy of the target driving trajectory.

[0013] According to another specific implementation of this application, a vehicle trajectory planning method is disclosed in this implementation, wherein the relationship between the vehicle's acceleration and velocity, the distance to the vehicle in front, and the desired velocity is determined by the following formula:

[0014]

[0015] in, To accelerate the vehicle, For speed, To maintain distance from the vehicle in front, For the desired speed, This is the vehicle's maximum acceleration. This represents the minimum expected safe distance for the vehicle.

[0016] According to another specific implementation of this application, the implementation of this application discloses a vehicle driving trajectory planning method. The learning-based trajectory planner includes an encoder and a multi-layer decoder. The vehicle trajectory prediction module determines multiple third predicted driving trajectories corresponding to the target vehicle based on third target information through the learning-based trajectory planner. This includes: using the encoder, determining the first feature information corresponding to the target obstacle and the second feature information corresponding to the map information based on the third target information, and obtaining target feature information based on the first and second feature information; and using the multi-layer decoder, determining multiple third predicted driving trajectories corresponding to the target vehicle based on the target feature information.

[0017] By employing the above technical solution, and utilizing an encoder and a multi-layer decoder, encoding and decoding processing is performed based on the third target information, resulting in more accurate multiple third predicted driving trajectories, thereby improving the accuracy of the target driving trajectory.

[0018] According to another specific implementation of this application, the vehicle trajectory planning method disclosed in this implementation method uses a multi-layer decoder to determine multiple third predicted driving trajectories corresponding to the target vehicle based on target feature information. The method includes: using the first layer decoder in the multi-layer decoder to generate a fourth predicted driving trajectory corresponding to the target vehicle and a fifth predicted driving trajectory corresponding to the target obstacle based on the target feature information, and sequentially inputting the fourth predicted driving trajectory and the fifth predicted driving trajectory generated by the previous layer decoder to the next layer decoder, until the last layer decoder generates the corresponding fourth predicted driving trajectory and the fifth predicted driving trajectory, and using the fourth predicted driving trajectory generated by the last layer decoder as the multiple third predicted driving trajectories corresponding to the target vehicle.

[0019] By adopting the above technical solution, using a staggered decoder to decode layer by layer, and employing a game theory-like approach, more accurate third-party predicted driving trajectories can be obtained, thereby improving the accuracy of the target driving trajectory.

[0020] According to another specific implementation of this application, the vehicle trajectory planning method disclosed in this implementation includes a deviation determination unit in the vehicle trajectory prediction module. The deviation determination unit includes a fully connected layer, which takes the fourth predicted driving trajectory generated by the last layer decoder as multiple third predicted driving trajectories corresponding to the target vehicle. The method includes: using the deviation determination unit to determine the deviation point corresponding to the fourth predicted driving trajectory based on the fourth predicted driving trajectory, and obtaining multiple third predicted driving trajectories based on the deviation point and the fourth predicted driving trajectory.

[0021] By adopting the above technical solution, the deviation point corresponding to the fourth predicted driving trajectory is obtained by using the deviation determination unit. Based on the deviation point and the fourth predicted driving trajectory, multiple more accurate third predicted driving trajectories can be obtained.

[0022] According to another specific implementation of this application, the implementation of this application discloses a vehicle trajectory planning method. The evaluation indicators include at least one of the following: collision without fault, compliance with drivable area, compliance with driving direction, vehicle progress relative to expert, collision time, ratio of planner progress along the route to expert progress, speed limit compliance, and comfort. The trajectory evaluation module determines the second evaluation results corresponding to each second predicted driving trajectory and each third predicted driving trajectory based on the first evaluation result and the evaluation indicators. This includes: selecting the first predicted driving trajectory corresponding to the first evaluation result that meets the first target condition as the first target predicted driving trajectory; determining the target evaluation indicator based on the first target predicted driving trajectory and the second predicted driving trajectory, or determining the target evaluation indicator based on the first target predicted driving trajectory and the third predicted driving trajectory; and determining the second evaluation results corresponding to each second predicted driving trajectory and each third predicted driving trajectory based on the target evaluation indicator.

[0023] By adopting the above technical solution, more accurate second evaluation results of each second predicted driving trajectory and each third predicted driving trajectory can be obtained, thereby improving the accuracy of the target driving trajectory.

[0024] According to another specific implementation of this application, a vehicle trajectory planning method is disclosed in this implementation, the method comprising determining a second evaluation result according to the following formula:

[0025]

[0026] in, This is the second evaluation result. To determine whether there is liability in the collision, For compliance with driving zone regulations, For compliance with driving direction, Based on the vehicle's progress relative to the expert, For collision time, The ratio of the planner's progress along the route to the expert's progress. For speed limit compliance, For comfort.

[0027] According to another specific implementation of this application, a vehicle trajectory planning method is disclosed in this implementation. The trajectory determination module determines the target trajectory information corresponding to the target vehicle based on the second evaluation result and the trajectory planning continuity requirement. This includes: determining the trajectory level corresponding to each second predicted trajectory and each third predicted trajectory based on the second evaluation result; determining at least one second target predicted trajectory based on the trajectory level, wherein the second target predicted trajectory is the trajectory with the highest trajectory level among the second predicted trajectory and each third predicted trajectory; obtaining the target trajectory determined at the previous moment during the target vehicle's driving process; if it is determined that at least one second target predicted trajectory includes a third target predicted trajectory that is the same as the target trajectory determined at the previous moment, then the third target predicted trajectory is determined as the target trajectory; if it is determined that at least one second target predicted trajectory does not include a third target predicted trajectory that is the same as the target trajectory determined at the previous moment, then a fourth target predicted trajectory that satisfies the second target condition is selected from at least one second target predicted trajectory as the target trajectory.

[0028] By adopting the above technical solution, the trajectory that connects most smoothly with the past vehicle trajectory can be selected from each of the second and third predicted driving trajectories as the target driving trajectory at the current moment, thereby improving the accuracy of the target driving trajectory and the comfort and safety of the vehicle during driving.

[0029] According to another specific implementation of this application, the vehicle driving trajectory planning method disclosed in this application further includes determining whether the second objective condition is met by: using a B-spline curve, predicting the driving trajectory based on the objective fitting window and the second objective, and determining the target fitting data point corresponding to the previous moment; if the error between the target fitting data point and the original data point corresponding to the target driving trajectory determined at the previous moment is less than the error threshold, it is determined that the objective condition is met.

[0030] By adopting the above technical solution, the trajectory that connects most smoothly with the past vehicle driving trajectory can be obtained as the target driving trajectory at the current moment, which improves the accuracy of the target driving trajectory, as well as the comfort and safety of the vehicle during driving.

[0031] Secondly, this application also discloses a vehicle trajectory planning method, characterized in that the method includes: determining multiple first predicted trajectories of target obstacles around a target vehicle based on first target information, wherein the first target information is information used for trajectory prediction, and the first target information includes historical trajectory information of the target vehicle, historical trajectory information of the target obstacles, and map information of the driving environment in which the target vehicle is located; determining multiple second predicted trajectories corresponding to the target vehicle using a rule-based trajectory planner based on second target information, and determining multiple third predicted trajectories corresponding to the target vehicle using a learning-based trajectory planner based on third target information; determining a first evaluation result corresponding to each first predicted trajectories, and determining a second evaluation result corresponding to each second predicted trajectories and each third predicted trajectories based on the first evaluation results and evaluation indicators; and determining the target vehicle's target trajectory based on the second evaluation results and trajectory planning continuity requirements.

[0032] Thirdly, this application also discloses a vehicle trajectory planning device, comprising: a first processing module, configured to determine multiple first predicted trajectories of target obstacles surrounding a target vehicle based on first target information; a second processing module, configured to determine multiple second predicted trajectories of the target vehicle based on second target information using a rule-based trajectory planner, and to determine multiple third predicted trajectories of the target vehicle based on third target information using a learning-based trajectory planner; a third processing module, configured to determine a first evaluation result corresponding to each first predicted trajectories, and to determine a second evaluation result corresponding to each second predicted trajectories and each third predicted trajectories based on the first evaluation results and evaluation indicators; and a fourth processing module, configured to determine the target vehicle's target trajectory based on the second evaluation results and trajectory planning continuity requirements.

[0033] Fourthly, this application also discloses an electronic device, including: a processor and a memory communicatively connected to the processor; the memory stores a computer program; the processor executes the computer program stored in the memory to enable the electronic device to implement the vehicle trajectory planning method provided by any of the implementations of the first or second aspect above.

[0034] Fifthly, this application provides a computer-readable storage medium storing a computer program, which, when executed by a processor, is used to implement the vehicle trajectory planning method provided by any of the first or second aspects described above.

[0035] Sixthly, an implementation of this application provides a computer program product, including a computer program that, when executed by a processor, implements the vehicle trajectory planning method provided by any of the implementations of the first or second aspect above.

[0036] It is understood that the beneficial effects of the second to sixth aspects mentioned above can also be found in the relevant descriptions in the first aspect mentioned above, and will not be repeated here. Attached Figure Description

[0037] Figure 1 This is a schematic diagram of a trajectory planning model provided in an embodiment of this application;

[0038] Figure 2 This is a flowchart illustrating a vehicle trajectory planning method provided in an embodiment of this application;

[0039] Figure 3 This is a flowchart illustrating a method for determining the second evaluation result corresponding to each second predicted driving trajectory and each third predicted driving trajectory, provided in an embodiment of this application.

[0040] Figure 4 This is a flowchart illustrating a method for determining the target driving trajectory information corresponding to a target vehicle, as provided in an embodiment of this application.

[0041] Figure 5 This is a flowchart illustrating a process for determining whether a second objective condition is met, provided in an embodiment of this application.

[0042] Figure 6 This is another flowchart illustrating the vehicle trajectory planning method provided in the embodiments of this application;

[0043] Figure 7 This is another flowchart illustrating the vehicle trajectory planning method provided in the embodiments of this application;

[0044] Figure 8This is a schematic diagram of a vehicle trajectory planning device provided in an embodiment of this application;

[0045] Figure 9 This is a schematic diagram of the structure of an electronic device provided in an embodiment of this application. Detailed Implementation

[0046] As mentioned earlier, in the prior art, the driving trajectory planning method for assisted driving is usually based on rule-based candidate trajectory prediction algorithms, and the vehicle's driving trajectory is obtained under the preset condition that other vehicles are all moving at a constant speed. The accuracy of the vehicle driving trajectory determined by this method is not stable, which has problems affecting vehicle driving efficiency, vehicle driving safety and user experience.

[0047] Based on this, this application provides a vehicle trajectory planning method. In the process of vehicle trajectory planning, based on the historical trajectory information of target obstacles around the vehicle, current operating status information, multiple first predicted trajectories, historical trajectory information of the target vehicle, current driving status information, and map information of the driving environment, multiple predicted trajectories of the target vehicle are generated from multiple aspects through rule-based trajectory planners and learning-based trajectory planners. Finally, based on the second evaluation results of the determined multiple predicted trajectories and the trajectory planning continuity requirements, the target vehicle trajectory planned from the multiple predicted trajectories is more accurate, effectively improving vehicle driving efficiency, driving safety, and user experience.

[0048] Next, with reference to the accompanying drawings, the steps and advantages of the vehicle trajectory planning method provided in this application will be described in detail.

[0049] In one implementation of this application, the vehicle trajectory planning method provided in this application is applied to, for example... Figure 1 The trajectory planning model shown includes an obstacle trajectory prediction module, a vehicle trajectory prediction module, a trajectory evaluation module, and a trajectory determination module. The vehicle trajectory prediction module includes a rule-based trajectory planner and a learning-based trajectory planner.

[0050] like Figure 2 As shown, the vehicle trajectory planning method provided in this application includes the following steps.

[0051] S100: The obstacle trajectory prediction module determines multiple first predicted driving trajectories of target obstacles around the target vehicle based on the first target information.

[0052] The first target information is used for trajectory prediction, and includes the historical trajectory information of the target vehicle, the historical trajectory information of the target obstacle, and the map information of the driving environment of the target vehicle.

[0053] In one implementation of this application, multiple first predicted driving trajectories are predicted driving trajectories for the next 8 seconds.

[0054] During the training of the trajectory planning model, the training dataset is prepared first. Specifically, the open-source autonomous driving planning dataset nuPlan is used, from which 63,837 scenarios are extracted and divided into training and validation sets in a 4:1 ratio. Each scenario consists of a 1-second historical trajectory and an 8-second future trajectory. Other relevant datasets can also be used.

[0055] After vectorizing the historical information in the scene and the surrounding map elements, they are transformed into tensor format for network input through padding and other methods. Through a multimodal trajectory prediction network architecture based on Transformer encoding-decoding, the predicted trajectory and corresponding score of each agent (i.e., an example of a target obstacle) around the vehicle are obtained. For each agent, the trajectory with the highest score is selected as the predicted trajectory of that agent.

[0056] S200: The vehicle trajectory prediction module determines multiple second predicted driving trajectories corresponding to the target vehicle based on the second target information and through a rule-based trajectory planner, and determines multiple third predicted driving trajectories corresponding to the target vehicle based on the third target information and through a learning-based trajectory planner.

[0057] The second target information includes the current driving status information of the target vehicle, and the third target information includes the current operating status information of the target obstacle.

[0058] S300: The trajectory evaluation module determines the first evaluation result corresponding to each first predicted driving trajectory, and determines the second evaluation result corresponding to each second predicted driving trajectory and each third predicted driving trajectory based on the first evaluation result and the evaluation index.

[0059] The first and second evaluation results can be evaluation scores, evaluation levels, etc.

[0060] S400: The trajectory determination module determines the target driving trajectory corresponding to the target vehicle based on the second evaluation result and the trajectory planning continuity requirements.

[0061] The vehicle trajectory planning method provided in this application includes the following steps: In the vehicle trajectory planning process, an obstacle trajectory prediction module determines multiple first predicted trajectories for target obstacles around the target vehicle based on first target information, including historical trajectory information of the target vehicle, historical trajectory information of the target obstacle, and map information of the target vehicle's driving environment. Then, a vehicle trajectory prediction module determines multiple second predicted trajectories for the target vehicle using a rule-based trajectory planner, based on second target information including the target vehicle's current driving state information. Finally, a learning-based trajectory planner determines multiple third predicted trajectories for the target vehicle based on map information, the second target information, and third target information including the current driving state information of the target obstacle. A trajectory evaluation module determines a first evaluation result for each first predicted trajectories and, based on the first evaluation results and evaluation indicators, determines a second evaluation result for each second and third predicted trajectories. Finally, a trajectory determination module determines the target vehicle's target trajectory based on the second evaluation results and the trajectory planning continuity requirements. Therefore, during the vehicle trajectory planning process, based on the historical trajectory information of surrounding obstacles, current operating status information, multiple first predicted trajectories, the historical trajectory information of the target vehicle, current driving status information, and map information of the driving environment, multiple predicted trajectories of the target vehicle are generated from multiple aspects through rule-based trajectory planners and learning-based trajectory planners. Finally, based on the second evaluation results of the determined multiple predicted trajectories and the trajectory planning continuity requirements, the target vehicle trajectory planned from the multiple predicted trajectories is more accurate, effectively improving vehicle driving efficiency, driving safety, and user experience.

[0062] In one implementation of this application, the first target information includes the historical trajectory information of the target vehicle, the historical trajectory information of the target obstacle, and the map information of the driving environment in which the target vehicle is located; the second target information includes the current driving state information of the target vehicle, which includes at least one of the following: acceleration information, speed information, distance information from the vehicle in front, desired speed information, and lane centerline information in the current driving environment; the third target information includes the map information of the driving environment in which the target vehicle is located, the current driving state information of the target vehicle, and the current motion state information of the target obstacle.

[0063] In one implementation of this application, the vehicle trajectory prediction module determines multiple second predicted driving trajectories corresponding to the target vehicle based on the second target information and through a rule-based trajectory planner, including: performing lateral trajectory planning processing and longitudinal trajectory planning processing on the second target information and through a rule-based trajectory planner to determine multiple second predicted driving trajectories corresponding to the target vehicle.

[0064] For example, rule-based planners (i.e., rule-based trajectory planners) use a combination of longitudinal and lateral planning methods from the Intelligent Driver Model (IDM) planning algorithm.

[0065] Among them, the intelligent driving model algorithm is mainly used for longitudinal planning and vehicle acceleration. Its speed Distance to the vehicle in front and expected speed The relationship between them is , where a represents the maximum acceleration of vehicle n. It is the expected minimum safe distance for vehicle n.

[0066] In lateral decision-making, three different offsets are selected based on the centerline. By setting the desired speed to 20%, 40%, 60%, 80%, and 100% of the road speed limit, 15 sets of candidate trajectories (i.e., second predicted driving trajectories) for different longitudinal and lateral programming are obtained.

[0067] In one implementation of this application, the learning-based trajectory planner includes an encoder and a multi-layer decoder. The vehicle trajectory prediction module determines multiple third predicted driving trajectories corresponding to the target vehicle based on the third target information through the learning-based trajectory planner. This includes: using the encoder to determine the first feature information corresponding to the target obstacle and the second feature information corresponding to the map information based on the third target information, and obtaining target feature information based on the first and second feature information; and using the multi-layer decoder to determine multiple third predicted driving trajectories corresponding to the target vehicle based on the target feature information.

[0068] In one implementation of this application, the vehicle trajectory prediction module uses a multi-layer decoder to determine multiple third predicted driving trajectories corresponding to the target vehicle based on target feature information. This includes: using the first layer decoder in the multi-layer decoder to generate a fourth predicted driving trajectory corresponding to the target vehicle and a fifth predicted driving trajectory corresponding to the target obstacle based on target feature information; and sequentially inputting the fourth and fifth predicted driving trajectories generated by the previous layer decoder into the next layer decoder until the last layer decoder generates the corresponding fourth and fifth predicted driving trajectories; and using the fourth predicted driving trajectory generated by the last layer decoder as the multiple third predicted driving trajectories corresponding to the target vehicle.

[0069] For example, the learning-based planner (i.e., the learning-based trajectory planner) employs a Transformer encoder-decoder architecture that simultaneously predicts the trajectories of other vehicles and its own vehicle. It is trained on a dataset based on imitation learning. The Transformer encoder uses different encoders for the agent and the map: AgentEncoder and MapEncoder, respectively, both simple fully connected layer networks. The extracted agent features (i.e., the first feature information) and map features (i.e., the second feature information) are concatenated after passing through the Transformer encoder. At the Transformer decoder level, a K-layer Transformer decoder (i.e., a multi-layer decoder) is used for layer-by-layer decoding, employing a game theory-like approach. The first layer of the Transformer decoder initially predicts the trajectory of each agent (i.e., the fifth predicted driving trajectory). Each subsequent layer updates its own trajectory based on the future trajectories of other vehicles obtained from the previous layer (i.e., the fourth predicted driving trajectory). Through this layer-by-layer game theory, the trajectory of each agent is finally obtained, with the vehicle's own trajectory used as a candidate trajectory (i.e., the third predicted driving trajectory).

[0070] In one implementation of this application, the vehicle trajectory prediction module further includes a deviation determination unit. The deviation determination unit includes a fully connected layer, which uses the fourth predicted driving trajectory generated by the last layer decoder as multiple third predicted driving trajectories corresponding to the target vehicle. The deviation determination unit determines the deviation point corresponding to the fourth predicted driving trajectory based on the fourth predicted driving trajectory, and obtains multiple third predicted driving trajectories based on the deviation point and the fourth predicted driving trajectory.

[0071] For example, a deviation network (i.e. deviation determination unit) is designed to further fine-tune the candidate trajectory (i.e. the fourth predicted driving trajectory) obtained by the intelligent driving model. The deviation network consists of simple fully connected layers. The candidate trajectories generated by the intelligent driving model will all pass through this deviation network to obtain the deviation of the trajectory points. The deviations are then summed to obtain the final candidate trajectory (i.e. the third predicted driving trajectory).

[0072] In one implementation of this application, the evaluation metrics include at least one of the following: no-fault collision, compliance with drivable area, compliance with driving direction, vehicle progress relative to expert, collision time, ratio of planner's progress along the route to expert's progress, speed limit compliance, and comfort. Figure 3 As shown, the trajectory evaluation module determines the second evaluation results corresponding to each second predicted driving trajectory and each third predicted driving trajectory based on the first evaluation result and the evaluation indicators, including the following steps.

[0073] S310: Select the first predicted driving trajectory corresponding to the first evaluation result that satisfies the first target condition as the first target predicted driving trajectory.

[0074] The first objective condition can be that the first evaluation result is optimal.

[0075] S320: Determine the target evaluation index based on the first target predicted driving trajectory and the second predicted driving trajectory, or determine the target evaluation index based on the first target predicted driving trajectory and the third predicted driving trajectory.

[0076] S330: Determine the second evaluation results corresponding to each second predicted driving trajectory and each third predicted driving trajectory based on the target evaluation indicators.

[0077] In one implementation of this application, the method includes determining a second evaluation result according to the following formula:

[0078]

[0079] in, This is the second evaluation result. To determine whether there is liability in the collision, For compliance with driving zone regulations, For compliance with driving direction, Based on the vehicle's progress relative to the expert, For collision time, The ratio of the planner's progress along the route to the expert's progress. For speed limit compliance, For comfort.

[0080] For example, based on the aforementioned 8-second trajectory prediction results (i.e., multiple first predicted driving trajectories), for each agent, the trajectory with the highest score (i.e., the first evaluation result) (i.e., satisfying the first objective condition) is selected as the predicted trajectory (i.e., the first objective predicted driving trajectory), thereby obtaining an occupancy map that can represent the space occupancy status at different future times.

[0081] For the aforementioned different candidate trajectories of the target vehicle (i.e., multiple second-predicted driving trajectories and multiple third-predicted driving trajectories), firstly, executable candidate trajectories are obtained based on simulation, and then these candidate trajectories are scored, including indicators such as whether or not a collision is at fault. Compliance with permitted driving areas Compliance of driving direction ,schedule (i.e., based on the vehicle's progress relative to the expert), collision time Self-progress (i.e., the ratio of the planner's progress along the route to the expert's progress), speed limit compliance. and comfort .

[0082] In one implementation of this application, This indicates that in the event of a collision, the responsibility for the collision will be determined based on the planner's fault; if the planner is at fault, the score is 0, otherwise the score is 1. Driving area compliance. This indicates that if the vehicle leaves the drivable area, the score is 0; otherwise, it is 1. Driving direction compliance. This indicates that if a vehicle travels in the wrong direction, the score is 0 or 0.5 based on the distance traveled in the wrong direction; otherwise, it is 1. Progress The score is calculated based on the vehicle's progress relative to the expert. If the planned path is less than 20% of the expert's progress, the score is 0; otherwise, the score is 1.

[0083] Collision time This indicates that the predicted vehicle and the detected trajectory are moving with lateral velocity and heading angle. TTC refers to the minimum time it takes for the vehicle to collide with the detected trajectory. A score of 1 is given if the TTC is greater than 0.95 seconds, otherwise 0. Self-progression This represents the ratio of the planner's progress along the route to the expert's progress. If the expert moves less than 0.1 meters, a low or negative progress score is ignored. Speed ​​limit compliance. Used to verify whether the planner is adhering to the speed limit of the current lane, taking into account both the duration and severity of the violation. Comfort. This is used to verify whether dynamic statistics such as acceleration are within a predefined threshold. A score of 1 is given if the threshold condition is met, and a score of 0 is given if the threshold condition is not met.

[0084] The final scoring module (i.e., the trajectory evaluation module) will calculate a final score for each candidate trajectory using the following formula: Ultimately, scores can be obtained for different candidate trajectories.

[0085] In one implementation of this application, such as Figure 4 As shown, the trajectory determination module determines the target driving trajectory information corresponding to the target vehicle based on the second evaluation result and the trajectory planning continuity requirements, including the following steps.

[0086] S410: Determine the trajectory level corresponding to each second predicted driving trajectory and each third predicted driving trajectory based on the second evaluation results.

[0087] S420: Determine at least one second target predicted driving trajectory based on the trajectory level.

[0088] Among them, the second target predicted driving trajectory is the driving trajectory with the highest trajectory level among the second predicted driving trajectory and each third predicted driving trajectory.

[0089] S430: Obtain the target vehicle's trajectory determined at the previous moment during the vehicle's operation.

[0090] S440: Determine whether at least one second target predicted driving trajectory includes a third target predicted driving trajectory that is the same as the target driving trajectory determined at the previous moment; if yes, proceed to step S450, otherwise proceed to step S460.

[0091] S450: Then the predicted driving trajectory of the third target is determined as the target driving trajectory.

[0092] S460: Then select the fourth target predicted driving trajectory that satisfies the second target condition from at least one second target predicted driving trajectory, and use it as the target driving trajectory.

[0093] For example, the scores for different trajectories can be divided into different levels (i.e., trajectory levels) according to different ranges, including a total of 5 levels: level 0, level 1, level 2, level 3, and level 4, used to control the switching of trajectories. The score (i.e., the second evaluation result) is set as level 0 for 0-0.2, level 1 for 0.2-0.4, level 2 for 0.4-0.6, level 3 for 0.6-0.8, and level 4 for 0.8-1.0.

[0094] The switching rule is based on whether the selected trajectories at the two consecutive time points are at the same level. Assume the trajectory selected at the previous time point t-1 is c (i.e., the target driving trajectory determined at the previous time point). Then, the following judgment is made: if trajectory c is still at the highest level among all candidate trajectories at time t, then that candidate trajectory is still selected; if trajectory c is not at the highest level at time t, then the trajectory with the smoothest connection to the past trajectory is selected from the highest level as the planned trajectory for that time point.

[0095] In one implementation of this application, such as Figure 5 As shown, the method also includes determining whether the second objective condition is met through the following steps.

[0096] S461: Using B-spline curves, determine the target fitting data points corresponding to the previous moment based on the target fitting window and the predicted driving trajectory of the second target.

[0097] S462: If the error between the target fitted data point and the original data point corresponding to the target driving trajectory determined in the previous moment is less than the error threshold, the second target condition is determined to be satisfied.

[0098] For example, a fitting window (i.e., the target fitting window) is selected around time t, and a curve is fitted using a B-spline curve. The difference (i.e., the error) between the fitted data points (i.e., the target fitted data points) and the original data points is calculated as a standard to measure whether the curve is smooth. The smaller the difference, the better the smoothness of the trajectory.

[0099] In another implementation of this application, the vehicle trajectory planning method provided in this application, such as... Figure 6 As shown, the steps include the following.

[0100] S10: Based on the first target information, determine multiple first predicted driving trajectories of target obstacles around the target vehicle.

[0101] The first target information is used for trajectory prediction, and includes the historical trajectory information of the target vehicle, the historical trajectory information of the target obstacle, and the map information of the driving environment of the target vehicle.

[0102] S20: Based on the second target information, a rule-based trajectory planner is used to determine multiple second predicted driving trajectories corresponding to the target vehicle, and based on the third target information, a learning-based trajectory planner is used to determine multiple third predicted driving trajectories corresponding to the target vehicle.

[0103] S30: Determine the first evaluation result corresponding to each first predicted driving trajectory, and determine the second evaluation result corresponding to each second predicted driving trajectory and each third predicted driving trajectory based on the first evaluation result and the evaluation index.

[0104] S40: Based on the second evaluation results and the trajectory planning continuity requirements, determine the target driving trajectory corresponding to the target vehicle.

[0105] In another implementation of this application, such as Figure 7 As shown, the vehicle trajectory planning method provided in this application includes the following steps.

[0106] By utilizing the historical trajectories of the vehicle itself and surrounding vehicles, as well as map information, this information is processed into the input of a neural network. The trajectory prediction network then obtains the predicted trajectories of other vehicles in the future (e.g., the next 8 seconds).

[0107] Different candidate trajectories are generated using different vehicle trajectory planners, including rule-based planners and learning-based planners, and the executable trajectory of the vehicle is obtained through simulation.

[0108] The scoring module scores different candidate trajectories generated by the planner based on trajectory prediction information and map information, and obtains scores for different trajectories.

[0109] By combining the scores of different candidate trajectories and taking into account the continuity of trajectory planning, the final execution trajectory (i.e. the final planning result) is selected from the candidate trajectories based on the planning results of the previous moment and handed over to the vehicle controller for execution to assist vehicle driving.

[0110] The vehicle trajectory planning method provided in this application, in the assisted driving planning task, uses a hybrid expert to generate candidate trajectories by simultaneously employing a learning-based planning algorithm (data-driven) and a rule-based planning algorithm (rule-driven), and selects the best candidate trajectory through simulation and scoring modules.

[0111] The vehicle trajectory planning method provided in this application is based on a hybrid expert planning framework, which integrates rule-based and learning-based planners. Combined with simulation verification and a refined scoring module, it effectively addresses the shortcomings in adaptability and generalization caused by the reliance on a single trajectory planning algorithm in assisted driving. Through the synergistic effect of a multimodal trajectory prediction network and trajectory scoring rules, it significantly improves the accuracy of trajectory prediction, the feasibility of planned trajectories, and the safety of driving behavior, achieving high accuracy in path planning, excellent risk control, and optimized driving comfort.

[0112] In another implementation of this application, such as Figure 8 As shown, a vehicle trajectory planning device is provided, comprising: a first processing module, configured to determine multiple first predicted trajectories of target obstacles surrounding a target vehicle based on first target information; a second processing module, configured to determine multiple second predicted trajectories of the target vehicle based on second target information using a rule-based trajectory planner, and to determine multiple third predicted trajectories of the target vehicle based on third target information using a learning-based trajectory planner; a third processing module, configured to determine a first evaluation result corresponding to each first predicted trajectories, and to determine a second evaluation result corresponding to each second predicted trajectories and each third predicted trajectories based on the first evaluation results and evaluation indicators; and a fourth processing module, configured to determine the target vehicle's target trajectory based on the second evaluation results and trajectory planning continuity requirements.

[0113] The first processing module is, for example, the obstacle trajectory prediction module mentioned above; the second processing module is, for example, the vehicle trajectory prediction module mentioned above; the third processing module is, for example, the trajectory evaluation module mentioned above; and the fourth processing module is, for example, the trajectory determination module mentioned above.

[0114] Please see Figure 9 , Figure 9 The diagram shown is a structural schematic of an electronic device provided in an embodiment of this application. Figure 9 As shown, the electronic device may include: transceiver 121, processor 122, and memory 123.

[0115] The processor 122 executes computer execution instructions stored in the memory, causing the processor 122 to execute part of the technical solution of the vehicle driving trajectory planning method in the above embodiments. The processor 122 can be a general-purpose processor, including a central processing unit (CPU), a network processor (NP), etc.; it can also be a digital data processor (DSP), an application-specific integrated circuit (ASIC), a field-programmable gate array (FPGA), or other programmable logic devices, discrete gate or transistor logic devices, or discrete hardware components.

[0116] The memory 123 is connected to the processor 122 via the system bus and completes communication between them. The memory 123 is used to store computer program instructions.

[0117] For example, and not as a limitation, memory 123 may include a hard disk drive (HDD), a floppy disk drive, flash memory, optical disk, magneto-optical disk, magnetic tape, or a universal serial bus (USB) drive, or a combination of two or more of these. Where appropriate, memory 123 may include removable or non-removable (or fixed) media. Where appropriate, memory 123 may be internal or external to the integrated gateway device. In a particular embodiment, memory 123 is non-volatile solid-state memory. In a particular embodiment, memory 123 includes read-only memory (ROM). Where appropriate, the ROM may be a mask-programmed ROM, a programmable ROM (PROM), an erasable PROM (EPROM), an electrically erasable PROM (EEPROM), an electrically rewritable ROM (EAROM), or flash memory, or a combination of two or more of these.

[0118] Transceiver 121 can be used to obtain the task to be run and its configuration information.

[0119] The system bus can be a Peripheral Component Interconnect (PCI) bus or an Extended Industry Standard Architecture (EISA) bus, etc. The system bus can be divided into address bus, data bus, control bus, etc. For ease of representation, only one thick line is used in the diagram, but this does not indicate that there is only one bus or one type of bus. Transceivers are used to enable communication between database access devices and other computers (e.g., clients, read-write libraries, and read-only libraries). Memory may include random access memory (RAM) and may also include non-volatile memory.

[0120] This application also provides a chip for executing instructions, which is used to execute the vehicle trajectory planning method described in the above embodiments.

[0121] This application also provides a computer-readable storage medium storing computer instructions. When these computer instructions are executed on the processor of an electronic device, the processor of the electronic device executes the vehicle trajectory planning method described in the above embodiments.

[0122] In some possible implementations, various aspects of the methods provided in this application can also be implemented as a program product, which includes program code. When the program product is run on the processor of an electronic device, the program code is used to cause the processor of the electronic device to perform the steps in the methods of the various exemplary implementations of this application described above. For example, the electronic device can execute the vehicle driving trajectory planning method described in the embodiments of this application.

[0123] The program product may employ any combination of one or more readable media. The readable media may be a readable data medium or a readable storage medium. A readable storage medium may be, for example, but not limited to, an electrical, magnetic, optical, electromagnetic, infrared, or semiconductor system, apparatus, or device, or any combination thereof. More specific examples of readable storage media (a non-exhaustive list) include: electrical connections with one or more wires, portable disks, hard disks, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM or flash memory), optical fibers, portable compact disk read-only memory (...). ( ), optical storage devices, magnetic storage devices, or any suitable combination thereof.

[0124] This application also provides a computer program product, which includes a computer program stored in a computer-readable storage medium. At least one processor can read the computer program from the computer-readable storage medium. When the at least one processor executes the computer program, it can implement the technical solution of the vehicle driving trajectory planning method in the above embodiments.

[0125] This application is described with reference to flowchart illustrations and / or block diagrams of the methods, apparatus, and computer program products according to this application. It should be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable information processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable information processing apparatus, generate instructions for implementing the flowchart... Figure 1 One or more processes and / or boxes Figure 1 A device that provides the functions specified in one or more boxes.

[0126] These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable information processing device to function in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means, which are implemented in a process Figure 1 One or more processes and / or boxes Figure 1 The function specified in one or more boxes.

[0127] These computer program instructions may also be loaded onto a computer or other programmable information processing equipment to cause a series of operational steps to be performed on the computer or other programmable equipment to produce a computer-implemented process, thereby providing instructions that execute on the computer or other programmable equipment for implementing the process. Figure 1 One or more processes and / or boxes Figure 1 The steps of the function specified in one or more boxes.

[0128] Other embodiments of this application will readily occur to those skilled in the art upon consideration of the specification and practice of the invention disclosed herein. This application is intended to cover any variations, uses, or adaptations of this application that follow the general principles of this application and include common knowledge or customary techniques in the art not disclosed herein.

[0129] It should be noted that, in addition to the specific embodiments described above, those skilled in the art can easily understand other advantages and effects of this application from the content disclosed in this specification. Although the description of this application is presented in conjunction with preferred embodiments, this does not mean that the features of this invention are limited to this implementation. On the contrary, the purpose of describing the invention in conjunction with the implementation is to cover other options or modifications that may be derived based on the claims of this application. To provide a thorough understanding of this application, many specific details are included in the above description, and this application may also be implemented without using these details. Furthermore, to avoid confusion or obscuring the focus of this application, some specific details will be omitted in the description. It should be noted that, unless otherwise specified, the embodiments and features in the embodiments of this application can be combined with each other.

[0130] It should be noted that in this specification, similar reference numerals and letters in the following figures indicate similar items. Therefore, once an item is defined in one figure, it does not need to be further defined and explained in subsequent figures.

[0131] Although this application has been illustrated and described with reference to certain preferred embodiments, those skilled in the art should understand that the above description is a further detailed explanation of the application in conjunction with specific implementations, and should not be construed as limiting the specific implementation of the application to these descriptions. Those skilled in the art can make various changes in form and detail, including some simple deductions or substitutions, without departing from the spirit and scope of this application.

Claims

1. A method for planning vehicle driving trajectory, characterized in that, The method is applied to a trajectory planning model, which includes an obstacle trajectory prediction module, a vehicle trajectory prediction module, a trajectory evaluation module, and a trajectory determination module. The vehicle trajectory prediction module includes a rule-based trajectory planner and a learning-based trajectory planner. The obstacle trajectory prediction module determines multiple first predicted driving trajectories of target obstacles around the target vehicle based on the first target information; The vehicle trajectory prediction module determines multiple second predicted driving trajectories corresponding to the target vehicle based on the second target information and through the rule-based trajectory planner, and determines multiple third predicted driving trajectories corresponding to the target vehicle based on the third target information and through the learning-based trajectory planner. The trajectory evaluation module determines the first evaluation result corresponding to each of the first predicted driving trajectories, and determines the second evaluation result corresponding to each of the second predicted driving trajectories and each of the third predicted driving trajectories based on the first evaluation result and the evaluation index. The trajectory determination module determines the target driving trajectory corresponding to the target vehicle based on the second evaluation result and the trajectory planning continuity requirement. Furthermore, the trajectory determination module determines the target driving trajectory information corresponding to the target vehicle based on the second evaluation result and the trajectory planning continuity requirement, including: Based on the second evaluation result, determine the trajectory level corresponding to each of the second predicted driving trajectories and each of the third predicted driving trajectories; At least one second target predicted driving trajectory is determined based on the trajectory level, and the second target predicted driving trajectory is the driving trajectory with the highest trajectory level among the second predicted driving trajectory and each of the third predicted driving trajectories. Obtain the target vehicle's trajectory determined at the previous moment during its driving process; If it is determined that the at least one second target predicted driving trajectory includes a third target predicted driving trajectory that is the same as the target driving trajectory determined at the previous moment, then the third target predicted driving trajectory is determined as the target driving trajectory; If it is determined that the at least one second target predicted driving trajectory does not include a third target predicted driving trajectory that is the same as the target driving trajectory determined at the previous moment, then a fourth target predicted driving trajectory that satisfies the second target condition is selected from the at least one second target predicted driving trajectory as the target driving trajectory.

2. The vehicle trajectory planning method as described in claim 1, characterized in that, The first target information includes the historical trajectory information of the target vehicle, the historical trajectory information of the target obstacle, and the map information of the driving environment of the target vehicle; The second target information includes the current driving status information of the target vehicle, which includes at least one of the following: acceleration information, speed information, distance information from the vehicle in front, desired speed information, and lane centerline information in the current driving environment. The third target information includes map information of the driving environment of the target vehicle, the current driving status information of the target vehicle, and the current motion status information of the target obstacle.

3. The vehicle trajectory planning method as described in claim 2, characterized in that, The vehicle trajectory prediction module, based on the second target information and through the rule-based trajectory planner, determines multiple second predicted driving trajectories corresponding to the target vehicle, including: Based on the second target information, the rule-based trajectory planner performs lateral trajectory planning and longitudinal trajectory planning to determine multiple second predicted driving trajectories corresponding to the target vehicle.

4. The vehicle trajectory planning method as described in claim 3, characterized in that, The learning-based trajectory planner includes an encoder and a multi-layer decoder. The vehicle trajectory prediction module, based on the third target information, determines multiple third predicted driving trajectories corresponding to the target vehicle through the learning-based trajectory planner, including: Using the encoder, based on the third target information, the first feature information corresponding to the target obstacle and the second feature information corresponding to the map information are determined, and the target feature information is obtained based on the first feature information and the second feature information; Using the multi-layer decoder, multiple third predicted driving trajectories corresponding to the target vehicle are determined based on the target feature information.

5. The vehicle trajectory planning method as described in claim 4, characterized in that, The vehicle trajectory prediction module utilizes the multi-layer decoder to determine multiple third predicted driving trajectories corresponding to the target vehicle based on the target feature information, including: Using the first layer of the multi-layer decoder, a fourth predicted driving trajectory corresponding to the target vehicle and a fifth predicted driving trajectory corresponding to the target obstacle are generated based on the target feature information. The fourth and fifth predicted driving trajectories generated by the previous layer decoder are then sequentially input into the next layer decoder until the last layer decoder generates the corresponding fourth and fifth predicted driving trajectories. The fourth predicted driving trajectory generated by the last layer decoder is then used as one of the multiple third predicted driving trajectories corresponding to the target vehicle.

6. The vehicle trajectory planning method as described in any one of claims 1-5, characterized in that, The evaluation metrics include at least one of the following: no-fault collision, compliance with drivable area, compliance with driving direction, vehicle progress relative to expert, collision time, ratio of planner progress along the route to expert progress, speed limit compliance, and comfort. The trajectory evaluation module determines the second evaluation result corresponding to each second predicted driving trajectory and each third predicted driving trajectory based on the first evaluation result and the evaluation metrics, including: The first predicted driving trajectory corresponding to the first evaluation result that satisfies the first objective condition is selected as the first objective predicted driving trajectory. The target evaluation index is determined based on the first target predicted driving trajectory and the second predicted driving trajectory, or the target evaluation index is determined based on the first target predicted driving trajectory and the third predicted driving trajectory. The second evaluation result corresponding to each of the second predicted driving trajectories and each of the third predicted driving trajectories is determined based on the target evaluation index.

7. The vehicle trajectory planning method as described in claim 6, characterized in that, The method includes determining the second evaluation result according to the following formula: in, This is the second evaluation result. Regarding whether or not there was a collision of responsibility, For the compliance of the drivable area, For the compliance of the driving direction, For the progress based on the vehicle relative to the expert, For the collision time, The ratio of the planner's progress along the route to the expert's progress. For the speed limit compliance, For the aforementioned comfort.

8. The vehicle trajectory planning method as described in claim 1, characterized in that, The method further includes determining whether the second objective condition is met by: Using B-spline curves, the target fitting data points corresponding to the previous moment are determined based on the target fitting window and the second target predicted driving trajectory. If the error between the target fitted data point and the original data point corresponding to the target driving trajectory determined at the previous moment is less than the error threshold, the second target condition is determined to be satisfied.

9. A method for planning vehicle driving trajectory, characterized in that, The method includes: Based on the first target information, determine multiple first predicted driving trajectories of target obstacles around the target vehicle; Based on the second target information, a rule-based trajectory planner determines multiple second predicted driving trajectories corresponding to the target vehicle, and based on the third target information, a learning-based trajectory planner determines multiple third predicted driving trajectories corresponding to the target vehicle. Determine the first evaluation result corresponding to each of the first predicted driving trajectories, and determine the second evaluation result corresponding to each of the second predicted driving trajectories and each of the third predicted driving trajectories based on the first evaluation result and the evaluation index; Based on the second evaluation result and the trajectory planning continuity requirement, the target driving trajectory corresponding to the target vehicle is determined, and the determination of the target driving trajectory information corresponding to the target vehicle based on the second evaluation result and the trajectory planning continuity requirement includes: Based on the second evaluation result, determine the trajectory level corresponding to each of the second predicted driving trajectories and each of the third predicted driving trajectories; At least one second target predicted driving trajectory is determined based on the trajectory level, and the second target predicted driving trajectory is the driving trajectory with the highest trajectory level among the second predicted driving trajectory and each of the third predicted driving trajectories. Obtain the target vehicle's trajectory determined at the previous moment during its driving process; If it is determined that the at least one second target predicted driving trajectory includes a third target predicted driving trajectory that is the same as the target driving trajectory determined at the previous moment, then the third target predicted driving trajectory is determined as the target driving trajectory; If it is determined that the at least one second target predicted driving trajectory does not include a third target predicted driving trajectory that is the same as the target driving trajectory determined at the previous moment, then a fourth target predicted driving trajectory that satisfies the second target condition is selected from the at least one second target predicted driving trajectory as the target driving trajectory.

10. An electronic device, characterized in that, include: A processor, and a memory communicatively connected to the processor; The memory stores computer programs; The processor executes the computer program stored in the memory to enable the electronic device to implement the vehicle trajectory planning method as described in any one of claims 1-8, or to enable the electronic device to implement the vehicle trajectory planning method as described in claim 9.

Images