A motion trajectory planning method, device and terminal equipment

Abstract

This application provides a motion trajectory planning method, apparatus, and terminal device, applicable to the field of data processing technology. The method includes: generating intermediate motion trajectory planning information based on initial motion trajectory planning information and preset safety area information; optimizing the intermediate motion trajectory planning information according to an intermediate motion trajectory optimization model to generate multiple intermediate motion trajectory optimization information; obtaining multiple candidate motion trajectory planning information based on the multiple intermediate motion trajectory optimization information, safety area information, and a motion trajectory prediction model; and filtering the multiple candidate motion trajectory planning information based on a motion trajectory filtering model to generate target motion trajectory planning information. This application combines safety area information to progressively optimize and generate a target trajectory that meets the requirements, ensuring the safety and rationality of vehicle driving, enabling vehicle driving to adapt to dynamic traffic scenarios, and improving the safety performance and operational reliability of the autonomous driving system.

Description

Technical Field

[0001] This application belongs to the field of data processing technology, and in particular relates to motion trajectory planning methods, devices and terminal equipment. Background Technology

[0002] In recent years, autonomous driving technology has made significant progress in core aspects such as perception, decision-making, and control, and is gradually expanding from closed scenarios to complex urban roads. The industry has placed higher demands on the safety, real-time performance, and generalization capabilities of autonomous vehicles. In particular, in dynamic traffic environments, how to achieve trajectory planning that balances safety, driving intent, and smoothness has become a key bottleneck for the technology's practical application.

[0003] Existing technologies are typically based on human-defined rules, generating trajectories through preset traffic regulations and obstacle avoidance logic, and relying on manually defined safety thresholds to assess risks.

[0004] However, current technologies cannot guarantee both safety and continuity of driving intent, easily leading to overly conservative deceleration or steering maneuvers that affect driving efficiency and passenger comfort. Furthermore, complex traffic scenarios significantly reduce the generalization ability of trajectory planning, making it difficult to adapt to diverse and complex real-world traffic conditions. Summary of the Invention

[0005] In view of this, embodiments of this application provide a motion trajectory planning method, apparatus, and terminal device, aiming to solve the problems in the prior art where safety and driving intention are difficult to coordinate, and the ability to adapt to complex traffic interaction behaviors is weak.

[0006] The first aspect of this application provides a motion trajectory planning method, including:

[0007] Obtain initial motion trajectory planning information;

[0008] Based on the initial motion trajectory planning information and the preset safety zone information, intermediate motion trajectory planning information is generated;

[0009] Based on the preset intermediate motion trajectory optimization model, the intermediate motion trajectory planning information is optimized to generate multiple intermediate motion trajectory optimization information.

[0010] Based on the multiple intermediate motion trajectory optimization information, the preset safety area information, and the preset motion trajectory prediction model, multiple candidate motion trajectory planning information are obtained.

[0011] Based on a preset motion trajectory screening model, the multiple candidate motion trajectory planning information is screened to generate target motion trajectory planning information.

[0012] A second aspect of this application provides a motion trajectory planning device, comprising:

[0013] The initial motion trajectory planning information acquisition module is used to acquire initial motion trajectory planning information;

[0014] The intermediate motion trajectory planning information generation module is used to generate intermediate motion trajectory planning information based on the initial motion trajectory planning information and the preset safety area information.

[0015] The intermediate motion trajectory optimization information generation module is used to optimize the intermediate motion trajectory planning information according to the preset intermediate motion trajectory optimization model, and generate multiple intermediate motion trajectory optimization information.

[0016] The candidate motion trajectory planning information generation module is used to obtain multiple candidate motion trajectory planning information based on the multiple intermediate motion trajectory optimization information, the preset safety area information, and the preset motion trajectory prediction model.

[0017] The target motion trajectory planning information generation module is used to filter the multiple candidate motion trajectory planning information based on a preset motion trajectory filtering model and generate target motion trajectory planning information.

[0018] A third aspect of this application provides a terminal device, the terminal device including a memory and a processor, the memory storing a computer program executable on the processor, the processor executing the computer program to implement the steps of the motion trajectory planning method as described in the first aspect above.

[0019] A fourth aspect of this application provides a computer-readable storage medium, comprising: storing a computer program, wherein when executed by a processor, the computer program implements the steps of the motion trajectory planning method described in the first aspect above.

[0020] Compared with the prior art, the beneficial effects of this application are: this application realizes online real-time safety assessment of the planned trajectory, with high efficiency, strong generalization and accurate safety assurance capabilities, can accurately identify and actively avoid potential collision risks, and gradually optimize and generate a target motion trajectory that meets the requirements by combining safe area information, so as to ensure the safety and rationality of vehicle driving, make vehicle driving adapt to dynamic traffic scenarios, and improve the safety performance and operational reliability of the autonomous driving system. Attached Figure Description

[0021] To more clearly illustrate the technical solutions in the embodiments of this application, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0022] Figure 1 This is a schematic diagram illustrating the implementation process of the motion trajectory planning method provided in Embodiment 1 of this application;

[0023] Figure 2 This is a schematic diagram illustrating the implementation process of the motion trajectory planning method provided in Embodiment 2 of this application;

[0024] Figure 3 This is a schematic diagram illustrating the implementation process of the motion trajectory planning method provided in Embodiment 3 of this application;

[0025] Figure 4 This is a schematic diagram illustrating the implementation process of the motion trajectory planning method provided in Embodiment 4 of this application;

[0026] Figure 5 This is a schematic diagram illustrating the implementation process of the motion trajectory planning method provided in Embodiment 5 of this application;

[0027] Figure 6 This is a schematic diagram illustrating the implementation process of the motion trajectory planning method provided in Embodiment Six of this application;

[0028] Figure 7 This is a schematic diagram illustrating the implementation process of the motion trajectory planning method provided in Embodiment 7 of this application;

[0029] Figure 8 This is a schematic diagram of the motion trajectory planning device provided in the embodiments of this application;

[0030] Figure 9 This is a schematic diagram of the terminal device provided in the embodiments of this application. Detailed Implementation

[0031] In the following description, specific details such as particular system architectures and techniques are set forth for illustrative purposes and not for limitation, in order to provide a thorough understanding of the embodiments of this application. However, those skilled in the art will understand that this application may also be implemented in other embodiments without these specific details. In other instances, detailed descriptions of well-known systems, apparatuses, circuits, and methods have been omitted so as not to obscure the description of this application with unnecessary detail.

[0032] To illustrate the technical solution described in this application, specific embodiments are provided below.

[0033] Figure 1A flowchart illustrating the implementation of the motion trajectory planning method provided in Embodiment 1 of this application is shown, and is described in detail below:

[0034] Step S101: Obtain initial motion trajectory planning information.

[0035] In this embodiment, the initial motion trajectory planning information may refer to the trajectory data initially generated by the autonomous vehicle in a specific traffic scenario based on its current position, target destination, and real-time road conditions to guide the vehicle's driving path. It may include key motion parameters such as the vehicle's position, speed, and acceleration over a period of time in the future, and may be the motion trajectory calculated by the autonomous vehicle's driving trajectory planner.

[0036] Step S102: Generate intermediate motion trajectory planning information based on the initial motion trajectory planning information and the preset safety area information.

[0037] In this embodiment, the preset safe area information can refer to a conflict-free occupancy area, used to represent the safe occupancy space of a vehicle within a given time interval. It is defined as a temporal set of grid occupancy states, which can consist of grid occupancy states at multiple times. Alternatively, it can be the input initial motion trajectory planning information, i.e., the initial planned trajectory. Areas that do not conflict with the preset safe zone information The initial planned trajectory This indicates that the vehicle is within the planning period (from time...) arrive The trajectory coordinate sequence at each moment, with no conflict-free occupied areas. Defined as a time-series set of grid occupancy states, the occupancy state of each grid point is determined by the formula. The calculation is performed, and if the grid point coordinates belong to the vehicle's predetermined safe driving area, the value is 1; otherwise, it is 0. In this way, an expert trajectory that meets the safety, smoothness, and continuity constraints can be generated using mixed integer programming as intermediate motion trajectory planning information.

[0038] Step S103: Based on the preset intermediate motion trajectory optimization model, optimize the intermediate motion trajectory planning information to generate multiple intermediate motion trajectory optimization information.

[0039] In this embodiment, the preset intermediate motion trajectory optimization model can be manually set, or it can be a mixed integer programming model. The optimization objective function can be defined as:

[0040] ;

[0041] ;

[0042] ;

[0043] in, For the first on the trajectory Line segment (trajectory point) arrive The length between (between) is used to constrain the overall length of the trajectory to be reasonable, preventing the trajectory from being too long or too short; For the first The smoothness loss term for each trajectory point is used to constrain the trajectory curvature, avoid excessive steering angles, and ensure the smoothness of the trajectory and the comfort of vehicle driving. and These are non-negative weighting coefficients used to adjust the relative importance between trajectory length and smoothness. In its constraint 1, the trajectory point constraint (position constraint), each trajectory point generated during optimization... Must be located within the vehicle's designated safe driving area Within this range, to ensure that trajectory points do not deviate from the safety boundary and always remain within a legal, conflict-free region. Its constraint 2, trajectory smoothness constraint (curvature constraint), It is the angle formed by three consecutive adjacent points on the trajectory, used to evaluate changes in trajectory curvature. Specifically, this angle... The calculation formula is:

[0044] ;

[0045] Among them, trajectory points , , Corresponding to the consecutive th in the trajectory sequence , , Coordinates of the trajectory points; vector dot product Describes the cosine value of the angle between two consecutive trajectory segments; The minimum included angle threshold (i.e., maximum permissible curvature) for the allowed trajectory is determined by limiting... To prevent sharp turns or large angle changes in the trajectory, thereby improving ride comfort and smoothness, a trajectory smoothness loss term is defined. for:

[0046] ;

[0047] Among them, the loss term is in the trajectory angle Less than the minimum allowable value Positive values ​​are generated to penalize trajectory segments that turn too sharply, ensuring the overall trajectory remains smooth. By processing the intermediate motion trajectory planning information using the aforementioned mixed-integer programming objective function and constraints, multiple intermediate motion trajectory optimization information that satisfy the constraints of safety, smoothness, and continuity can be generated; these are known as intermediate motion trajectory optimization information.

[0048] Step S104: Based on the multiple intermediate motion trajectory optimization information, the preset safety area information, and the preset motion trajectory prediction model, multiple candidate motion trajectory planning information are obtained.

[0049] In this embodiment, the preset motion trajectory prediction model can be a trained Transformer network based on a multi-head self-attention mechanism. It can be generated by using multiple intermediate motion trajectory optimization information and preset safety region information as input data, and then processing these as calculations to produce multiple candidate motion trajectory planning information. Alternatively, it can first receive a grid state sequence as input and encode it as high-dimensional features for subsequent trajectory prediction. Input feature definition: The network input sequence is represented as:

[0050] ;

[0051] in, The length of the grid sequence is the total number of spatial grid points involved in trajectory prediction. For each grid point, there are state feature dimensions, such as occupancy status, velocity information, or other environmental features.

[0052] Attention mechanism feature calculation: The input sequence first passes through a fully connected network (Multilayer Perceptron MLP, denoted as...) The feature transformation is performed to generate three feature matrices—query matrices—used for attention calculation. Key matrix AND-value matrix :

[0053] ;

[0054] in, This indicates that the input features are processed through a fully connected layer. By performing mapping, a high-dimensional hidden feature representation is obtained. , , These are learnable weight matrices used to further map hidden features to query, key, and value spaces.

[0055] Attention weight calculation: based on the above generated The matrix is ​​used to calculate attention weights using the Scaled Dot-Product method to obtain the network's ability to capture the correlations between environmental features:

[0056] ;

[0057] in, For query matrix AND key matrix The product of features is used to calculate the similarity or correlation between features; The dimension of the key feature is used to scale the dot product result to avoid gradient vanishing when the dimension is too large. The function normalizes the obtained dot product weights to obtain the attention weight distribution, and then multiplies it by the value matrix. Obtain the fused features .

[0058] Output feature computation: Finally, the attention features are processed through another nonlinear transformation layer (Multilayer Perceptron, MLP). The process is performed to output the final feature vector used to predict the trajectory coordinates. :

[0059] ;

[0060] in, The output features used to predict safe trajectories for the Transformer network will be further mapped to the trajectory point coordinate space for the generation of optimized safe trajectories.

[0061] In this embodiment, during the Transformer network training phase, this patent employs a supervised learning method, defining the loss function as the distance between the network's output trajectory points and the expert trajectory points, so that the network's predicted trajectory is as close as possible to the expert's demonstration trajectory.

[0062] ;

[0063] in, The parameters of the Transformer network to be optimized; Represents the coordinates of the trajectory points predicted by the Transformer network; Represents the true coordinates of the expert's trajectory; Indicates the distance norm (e.g., L2 norm); The weighting coefficients, used to control the intensity of the loss, are adjusted to regulate the penalty for prediction errors. By minimizing the above loss function, the Transformer network can efficiently learn expert trajectory patterns, thereby obtaining an initial trajectory prediction strategy with safe heuristics.

[0064] Step S105: Based on the preset motion trajectory screening model, the multiple candidate motion trajectory planning information is screened to generate target motion trajectory planning information.

[0065] In this embodiment, the preset motion trajectory selection model can be manually set, a reinforcement learning model, or it can be obtained by using multiple candidate motion trajectory planning information as input data and then calculating the target motion trajectory planning information through the motion trajectory selection model. Alternatively, the interaction between the vehicle and the environment can be modeled as a Markov decision process. ,in, State space represents the set of various environmental states in which the autonomous vehicle is located; The action space refers to the state of the vehicle. The set of possible actions that can be taken; Let be the state transition probability function, defined as the state transition probability function. Execute action Then transition to the next state The probability, i.e. ; This is the reward function, used to measure the quality of vehicle behavior to guide policy optimization; Let be the initial state distribution, representing the probability distribution of the initial state when the vehicle begins making a decision; This is a discount factor used to balance the importance of current rewards and future rewards.

[0066] In this embodiment, the policy network and value network of the reinforcement learning model can be trained using the PPO (Proximal Policy Optimization) algorithm. The policy network outputs the selection probability distribution of each candidate trajectory planning information based on the Transformer model, while the value network estimates the value of the current state. The selection policy is iteratively optimized by alternately updating the policy network (increasing the probability of selecting high-reward candidate trajectories) and the value network (optimizing the state value estimation). When training converges, for multiple input candidate trajectory planning information, the policy network outputs their corresponding probabilities and selects the candidate trajectory planning information with the highest probability as the target trajectory planning information.

[0067] The motion trajectory planning method provided in this application realizes online real-time safety assessment of the planned trajectory, and has high efficiency, strong generalization and accurate safety assurance capabilities. It can accurately identify and actively avoid potential collision risks, and combine safe area information to gradually optimize and generate a target motion trajectory that meets the requirements, so as to ensure the safety and rationality of vehicle driving, make vehicle driving adapt to dynamic traffic scenarios, and improve the safety performance and operational reliability of the autonomous driving system.

[0068] Figure 2 The flowchart illustrating the motion trajectory planning method provided in Embodiment 2 of this application is shown. The difference between this method and Embodiment 1 is that step S102 specifically includes:

[0069] Step S201: Based on the initial motion trajectory planning information, the time information and position coordinate information are processed to generate the initial motion trajectory time coordinate information.

[0070] In this embodiment, the initial motion trajectory planning information includes key motion parameters such as the vehicle's position, speed, and acceleration over a future period of time. These parameters are organized in chronological order (from the start time to the end time of the planning cycle), and the position coordinates corresponding to each moment are extracted to form a coordinate sequence arranged in chronological order, which yields the initial motion trajectory time-series coordinate information. This information can clearly reflect the expected position changes of the vehicle at different times.

[0071] Step S202: Generate safe area time-series coordinate information based on the preset safe area information, including time information and location coordinate information.

[0072] In this embodiment, the preset safe area information is a conflict-free occupied area, which is a time-series set of grid occupancy states. The grid occupancy state at each moment is analyzed, and the position coordinates of all grid points in the safe driving area at each moment are extracted. These coordinates are arranged in chronological order to form a coordinate sequence of the safe area at different moments, that is, to generate the temporal coordinate information of the safe area. This information can reflect the spatial changes of the safe area over time.

[0073] Step S203: Generate safe area position coordinate occupancy information based on the initial motion trajectory time coordinate information and the safe area time coordinate information.

[0074] In this embodiment, the position coordinates of each moment in the initial motion trajectory time-series coordinate information are compared with the position coordinates of the safe area at the corresponding moment in the safe area time-series coordinate information to determine whether each coordinate point on the initial motion trajectory is located within the safe area at the corresponding moment. If it is located within the safe area, the safe area position coordinate occupancy information corresponding to that coordinate point is 1; otherwise, it is 0. This generates the safe area position coordinate occupancy information, which clarifies the positional relationship between the initial motion trajectory and the safe area.

[0075] Step S204: Based on the location coordinate occupancy information of the safe area and the preset safe area information, the initial motion trajectory planning information is smoothed to generate intermediate motion trajectory planning information.

[0076] In this embodiment, by combining the location coordinate occupancy information of the safe area, the coordinate points located outside the safe area in the initial motion trajectory planning information are identified. Based on the preset safe area information (i.e., the non-conflicting occupied area), the positions of these coordinate points are adjusted so that the adjusted trajectory points are all within the safe area. At the same time, the trajectory is corrected by using smoothness constraints (such as limiting the included angle formed by three consecutive adjacent points to not less than the minimum included angle threshold) to ensure the smoothness and continuity of the trajectory, and finally generating intermediate motion trajectory planning information.

[0077] The motion trajectory planning method provided in this application performs time-series processing and coordinate comparison of initial motion trajectory planning information and preset safety area information, and meticulously analyzes the spatiotemporal relationship between the two. This allows the generated intermediate motion trajectory planning information to more accurately meet safety requirements, further improving the safety and reliability of motion trajectory planning. This enables autonomous vehicles to adapt more flexibly to changes in the safety area in dynamic traffic environments, ensuring a smooth and safe driving process.

[0078] Figure 3 The flowchart illustrating the motion trajectory planning method provided in Embodiment 3 of this application is shown. The difference between this method and Embodiment 1 is that step S103 specifically includes:

[0079] Step S301: Extract coordinate position information based on the intermediate motion trajectory planning information to generate multiple intermediate motion trajectory coordinate position information and multiple intermediate motion trajectory line segment information.

[0080] In this embodiment, the intermediate motion trajectory planning information is an expert trajectory that meets the constraints of safety, smoothness and continuity. The coordinate point data corresponding to each moment is extracted from the trajectory to obtain multiple intermediate motion trajectory coordinate position information. At the same time, two adjacent coordinate points are connected in sequence to form line segments. Each line segment corresponds to a path on the trajectory, thereby generating multiple intermediate motion trajectory line segment information. These information can reflect the specific location points and line segment composition of the trajectory respectively.

[0081] Step S302: Calculate the intermediate motion curvature angle information based on the coordinate position information of the multiple intermediate motion trajectories and the line segment information of the multiple intermediate motion trajectories.

[0082] In this embodiment, the coordinate position information of multiple intermediate motion trajectories is arranged in chronological order. Three consecutive adjacent coordinate points are selected (such as the (i-2), (i-1), and (i)th coordinate points). Based on the intermediate motion trajectory line segment information, the corresponding line segments are determined (the line segment from the (i-2)th point to the (i-1)th point, and the line segment from the (i-1)th point to the (i)th point). By calculating the angle between these two line segments, the intermediate motion curvature angle information is obtained. This information is used to evaluate the curvature change of the trajectory at this position.

[0083] Step S303: Determine whether the intermediate motion curvature angle information is less than the preset intermediate motion curvature angle threshold; if yes, proceed to step S304; if no, generate multiple intermediate motion trajectory optimization information based on the intermediate motion trajectory planning information.

[0084] In this embodiment, the preset intermediate motion curvature angle threshold can be manually set, and may refer to the minimum angle threshold, used to limit the maximum allowable curvature of the trajectory. The calculated intermediate motion curvature angle information is compared with this threshold. If the intermediate motion curvature angle information is less than the preset intermediate motion curvature angle threshold, it indicates that the trajectory has a risk of sharp turning at this position and needs to be optimized. If it is not less than the intermediate motion curvature angle threshold, it indicates that the current curvature of the trajectory meets the smoothness requirements and no additional optimization is needed. Multiple intermediate motion trajectory optimization information that meet the constraints of safety, smoothness, and continuity can be directly generated based on the intermediate motion trajectory planning information.

[0085] Step S304: Based on the preset intermediate motion trajectory optimization model, optimize the coordinate position information of the multiple intermediate motion trajectories to generate multiple optimized coordinate position information of intermediate motion trajectories.

[0086] In this embodiment, the preset intermediate motion trajectory optimization model can be a mixed integer programming model. When the intermediate motion curvature angle information is less than a preset threshold, the optimization objective function of mixed integer programming (minimizing the weighted sum of trajectory length and smoothness loss term) is used as a guide to adjust the coordinate position information of multiple intermediate motion trajectories. By changing the position of the coordinate points, the angle formed by three consecutive adjacent points after adjustment is not less than the preset intermediate motion curvature angle threshold, thereby generating multiple optimized coordinate position information of intermediate motion trajectories to ensure the smoothness of the trajectory.

[0087] Step S305: Generate multiple intermediate motion trajectory optimization information based on the multiple intermediate motion trajectory optimization coordinate position information.

[0088] In this embodiment, the optimized coordinate position information of multiple intermediate motion trajectories is arranged in chronological order to form a complete trajectory sequence. Each sequence corresponds to an optimized trajectory, that is, multiple intermediate motion trajectory optimization information is generated, so that the generated motion trajectory satisfies both the safety area constraint and the smoothness requirement.

[0089] The motion trajectory planning method provided in this application first extracts coordinate and line segment information, calculates the curvature angle and compares it with a threshold, and then optimizes the trajectory part that does not meet the smoothness requirements in a targeted manner. This improves the accuracy and reliability of the intermediate motion trajectory optimization information, so that the generated trajectory can better balance safety and driving comfort in complex traffic scenarios, thereby enhancing the trajectory planning performance of autonomous vehicles.

[0090] Figure 4 The flowchart illustrating the motion trajectory planning method provided in Embodiment 4 of this application is shown. The difference between this method and Embodiment 1 is that step S104 specifically includes:

[0091] Step S401: Obtain the spatial coordinate information of the predicted motion trajectory.

[0092] In this embodiment, the motion trajectory prediction spatial location coordinate information refers to the set of coordinates corresponding to the spatial range that the autonomous vehicle may travel within the planning period. This range covers all the locations that the vehicle may reach in the future and includes the location coordinates of multiple grid points. These coordinates together constitute the spatial region used for trajectory prediction.

[0093] Step S402: Based on the position coordinate information of the multiple intermediate motion trajectory optimization information and the motion trajectory prediction spatial position coordinate information, generate multiple motion trajectory prediction spatial position coordinate occupancy status information.

[0094] In this embodiment, multiple intermediate motion trajectory optimization information includes their respective position coordinate information. The position coordinates of each intermediate motion trajectory optimization information are compared with the grid point coordinates in the motion trajectory prediction space position coordinate information to determine whether the coordinates of the intermediate motion trajectory optimization information are located within the motion trajectory prediction space. If they are located within the space, the corresponding motion trajectory prediction space position coordinate occupancy status information is 1; otherwise, it is 0. This generates multiple motion trajectory prediction space position coordinate occupancy status information.

[0095] Step S403: Based on the spatial location coordinate occupancy status information of the multiple motion trajectory predictions and the preset motion trajectory prediction model, multiple candidate motion trajectory planning information is obtained.

[0096] In this embodiment, the preset motion trajectory prediction model is a Transformer network based on a multi-head self-attention mechanism. The occupancy status information of multiple motion trajectory prediction spatial position coordinates is used as the input of the network. This information is transformed through a fully connected network to generate a query matrix, a key matrix, and a value matrix. The fused features are obtained by calculating attention weights and then processed through a nonlinear transformation layer to output feature vectors for predicting trajectory coordinate points. Finally, these vectors are mapped to the trajectory point coordinate space to obtain multiple candidate motion trajectory planning information.

[0097] The motion trajectory planning method provided in this application clarifies the spatial range of motion trajectory prediction and generates corresponding occupancy status information, enabling the preset motion trajectory prediction model to more accurately combine intermediate motion trajectory optimization information for trajectory prediction. This makes the generated candidate motion trajectory planning information more in line with the spatial constraints of actual driving, further improving the accuracy and applicability of trajectory prediction and enhancing the adaptability of the autonomous driving system to dynamic traffic scenarios.

[0098] Figure 5 The flowchart illustrating the motion trajectory planning method provided in Embodiment 5 of this application is shown. The difference between this method and Embodiment 1 is that step S105 specifically includes:

[0099] Step S501: Based on the multiple candidate motion trajectory planning information, the preset safe area indication function, and the preset motion trajectory safety penalty coefficient, the motion trajectory safety characterization information is calculated.

[0100] In this embodiment, a preset safe zone indicator function is used to determine whether the coordinates of the candidate motion trajectory planning information are located within the preset safe zone information. If they are within the safe zone, the function value is 1; otherwise, it is 0. The preset motion trajectory safety penalty coefficient is a non-negative coefficient used to adjust the penalty intensity for unsafe trajectories. The coordinates of each trajectory point in multiple candidate motion trajectory planning information are substituted into the safe zone indicator function, and combined with the motion trajectory safety penalty coefficient, a penalty calculation is performed on trajectory points located outside the safe zone. The penalty results of all trajectory points are summarized to obtain the motion trajectory safety characterization information corresponding to each candidate motion trajectory planning information. This information is used to quantitatively evaluate the safety level of the trajectory.

[0101] Step S502: Based on the preset motion trajectory screening model, the multiple candidate motion trajectory planning information is screened according to the motion trajectory quality characterization information to generate target motion trajectory planning information.

[0102] In this embodiment, the preset motion trajectory screening model is a reinforcement learning model. The motion trajectory quality representation information includes motion trajectory safety representation information, and may also include trajectory smoothness and consistency with the initial intention. Figure 1 Evaluation metrics such as consistency are used. Based on these quality representations, the policy network of the reinforcement learning model evaluates multiple candidate trajectory planning information and outputs their respective selection probability distributions. The value network estimates the value of choosing different trajectories in the current state. Through iterative optimization, the policy network tends to choose the trajectory with better quality. Finally, the candidate trajectory planning information with the highest probability is selected as the target trajectory planning information.

[0103] The motion trajectory planning method provided in this application introduces motion trajectory safety representation information and quality representation information, enabling the reinforcement learning model to more comprehensively consider safety and other key indicators when selecting candidate trajectories. This further improves the reliability and applicability of the target motion trajectory planning information, ensuring that autonomous vehicles can guarantee safety while also taking into account driving smoothness and continuity of intent in complex traffic environments.

[0104] Figure 6 The flowchart illustrating the motion trajectory planning method provided in Embodiment Six of this application is shown. The difference between this method and Embodiment Five is that step S501 specifically includes:

[0105] Step S601: Based on the multiple candidate motion trajectory planning information, the preset motion trajectory safety penalty coefficient, the preset motion trajectory smoothness penalty coefficient, and the preset motion trajectory deviation coefficient, the motion trajectory quality characterization information is calculated.

[0106] In this embodiment, a preset motion trajectory safety penalty coefficient is used to penalize trajectory points located outside the safe zone in the candidate motion trajectory planning information; a preset motion trajectory smoothness penalty coefficient is used to penalize segments with excessive curvature in the trajectory; and a preset motion trajectory deviation coefficient is used to measure the degree of deviation between the candidate trajectory and the initial motion trajectory planning information. By multiplying these coefficients by their corresponding evaluation indicators and then summing them, the motion trajectory quality characterization information of each candidate motion trajectory planning information is comprehensively calculated. This information fully reflects the safety, smoothness, and consistency with the initial intent of the trajectory. The evaluation indicators include the degree of safety violation, the degree of curvature exceeding the standard, and the deviation distance.

[0107] Step S602: Extract coordinate position information from the multiple candidate motion trajectory planning information to generate multiple candidate motion trajectory coordinate position information and multiple candidate motion trajectory line segment information.

[0108] In this embodiment, the multiple candidate motion trajectory planning information is trajectory data generated by the Transformer network. The coordinate points corresponding to each time moment are extracted from each candidate trajectory to obtain the coordinate position information of multiple candidate motion trajectories. At the same time, adjacent coordinate points are connected in sequence to form line segments. Each line segment corresponds to a part of the trajectory path, thereby generating multiple candidate motion trajectory line segment information. This information reflects the specific location points and line segment composition of the trajectory.

[0109] Step S603: Calculate the curvature angle information of multiple candidate motion trajectories based on the coordinate position information and line segment information of the multiple candidate motion trajectories.

[0110] In this embodiment, the coordinate position information of multiple candidate motion trajectories is arranged in chronological order. Three consecutive adjacent coordinate points are selected (such as the (i-2), (i-1), and (i)th coordinate points). Based on the candidate motion trajectory line segment information, the corresponding line segments are determined (the line segment from the (i-2)th point to the (i-1)th point, and the line segment from the (i-1)th point to the (i)th point). By calculating the included angle between these two line segments, the curvature angle information of each candidate trajectory at the corresponding position is obtained. This information is used to evaluate the smoothness of the trajectory.

[0111] Step S604: Based on the curvature angle information of the multiple candidate motion trajectories and the preset motion trajectory smoothness penalty coefficient, the motion trajectory smoothness characterization information is calculated.

[0112] In this embodiment, the preset motion trajectory smoothness penalty coefficient can be manually set and can be a non-negative coefficient, used to adjust the penalty intensity for trajectories with excessive curvature. The curvature angle information of multiple candidate motion trajectories is compared with a preset minimum angle threshold. For angles smaller than the threshold, a penalty value is calculated based on the motion trajectory smoothness penalty coefficient. All penalty values ​​are then summarized to obtain the motion trajectory smoothness characterization information for each candidate trajectory. This information quantitatively reflects the smoothness of the trajectory.

[0113] Step S605: Extract coordinate position information from the initial motion trajectory planning information to generate multiple initial motion trajectory coordinate position information.

[0114] In this embodiment, the initial motion trajectory planning information includes the vehicle's position coordinate sequence within the planning period. The coordinate point data corresponding to each moment is extracted from the trajectory to generate multiple initial motion trajectory coordinate position information, which reflects the vehicle's initially planned path position.

[0115] Step S606: Calculate the distance between the coordinate position information of the multiple candidate motion trajectories and the coordinate position information of the multiple initial motion trajectories to obtain the motion trajectory coordinate position offset information.

[0116] In this embodiment, multiple candidate trajectory coordinate position information and multiple initial trajectory coordinate position information are arranged in chronological order. The distance (such as Euclidean distance) between the candidate trajectory coordinate point and the initial trajectory coordinate point at the same moment is calculated to obtain the position offset value at each moment. The offset values ​​at all moments are summarized to form the trajectory coordinate position offset information, which reflects the degree of deviation between the candidate trajectory and the initial trajectory.

[0117] Step S607: Based on the motion trajectory coordinate position offset information and the preset motion trajectory offset degree coefficient, calculate the motion trajectory offset degree characterization information.

[0118] In this embodiment, the preset trajectory deviation coefficient can be manually set and can be a non-negative coefficient, used to adjust the penalty for trajectory deviation. The deviation value at each moment in the trajectory coordinate position deviation information is multiplied by this coefficient and then summed to obtain the trajectory deviation characterization information. This information quantitatively reflects the overall deviation between the candidate trajectory and the initial trajectory planning information.

[0119] Step S608: Based on the motion trajectory safety characterization information, motion trajectory smoothness characterization information, and motion trajectory deviation characterization information, the motion trajectory quality characterization information is calculated.

[0120] In this embodiment, the motion trajectory safety representation information, motion trajectory smoothness representation information, and motion trajectory deviation degree representation information are respectively based on safety, smoothness, and intention. Figure 1 The trajectory quality is evaluated from three dimensions: consistency, ...

[0121] The motion trajectory planning method provided in this application provides quantitative evaluation and comprehensive integration from three dimensions: safety, smoothness, and deviation. This enables the motion trajectory screening model to more accurately identify the optimal trajectory, further improving the rationality and applicability of the target motion trajectory planning information, and ensuring that autonomous vehicles can achieve safe, smooth, and consistent driving in complex traffic environments.

[0122] Figure 7 The flowchart illustrating the motion trajectory planning method provided in Embodiment Seven of this application is shown. The difference between this method and Embodiment Five is that step S502 specifically includes:

[0123] Step S701: Optimize the preset motion trajectory screening model according to the preset motion trajectory screening model to obtain the optimized motion trajectory screening model.

[0124] In this embodiment, the preset motion trajectory selection model can be a reinforcement learning model, and the preset motion trajectory selection model optimization model can be a proximal policy optimization algorithm model, i.e., the PPO algorithm model. Sample data of vehicle-environment interaction (including state, action, reward, and next state) can be collected. The pruning objective function of the PPO algorithm is used to limit the policy update magnitude, and the policy network and value network are updated alternately: the policy network outputs the selection probability of candidate trajectories based on motion trajectory quality representation information, optimizing the policy by maximizing the cumulative reward; the value network optimizes the state value estimation by minimizing the error between the predicted value and the actual cumulative reward. After multiple rounds of iterative training, the optimized motion trajectory selection model is obtained, improving the model's ability to identify high-quality trajectories.

[0125] Step S702: Based on the optimized motion trajectory screening model, the multiple candidate motion trajectory planning information is screened according to the motion trajectory quality characterization information to generate target motion trajectory planning information.

[0126] In this embodiment, the optimized motion trajectory selection model receives motion trajectory quality representation information (including dimensions such as safety, smoothness, and deviation). The policy network evaluates multiple candidate motion trajectory planning information and outputs their respective selection probability distributions, selecting the candidate trajectory with the highest probability as the target motion trajectory planning information. During this process, the value network assists in evaluating the long-term benefits of different trajectory selections, ensuring that the selected target trajectory can meet safety requirements in a dynamic traffic environment while also taking into account driving comfort and consistency with the initial intent.

[0127] The motion trajectory planning method provided in this application enhances the model's ability to identify and select high-quality trajectories in complex traffic scenarios by specifically optimizing the motion trajectory screening model. This makes the generated target motion trajectory planning information more in line with actual driving needs, further improving the trajectory planning accuracy, safety, and reliability of autonomous vehicles.

[0128] Corresponding to the method in the above embodiments, Figure 8 A structural block diagram of the motion trajectory planning device provided in the embodiments of this application is shown. For ease of explanation, only the parts related to the embodiments of this application are shown. Figure 8 The example motion trajectory planning device can be the execution subject of the motion trajectory planning method provided in the aforementioned embodiment 1.

[0129] Reference Figure 8 The motion trajectory planning device includes:

[0130] The initial motion trajectory planning information acquisition module 810 is used to acquire initial motion trajectory planning information;

[0131] The intermediate motion trajectory planning information generation module 820 is used to generate intermediate motion trajectory planning information based on the initial motion trajectory planning information and the preset safety area information.

[0132] The intermediate motion trajectory optimization information generation module 830 is used to optimize the intermediate motion trajectory planning information according to the preset intermediate motion trajectory optimization model, and generate multiple intermediate motion trajectory optimization information.

[0133] The candidate motion trajectory planning information generation module 840 is used to obtain multiple candidate motion trajectory planning information based on the multiple intermediate motion trajectory optimization information, the preset safe area information and the preset motion trajectory prediction model.

[0134] The target motion trajectory planning information generation module 850 is used to filter the multiple candidate motion trajectory planning information based on a preset motion trajectory filtering model to generate target motion trajectory planning information.

[0135] The process by which each module in the motion trajectory planning device provided in this application implements its respective function can be specifically referred to the foregoing. Figure 1 The description of Embodiment 1 shown will not be repeated here.

[0136] It should be understood that the sequence number of each step in the above embodiments does not imply the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of this application.

[0137] It should be understood that, when used in this application specification and the appended claims, the term "comprising" indicates the presence of the described features, integrals, steps, operations, elements and / or components, but does not exclude the presence or addition of one or more other features, integrals, steps, operations, elements, components and / or a collection thereof.

[0138] It should also be understood that the term “and / or” as used in this application specification and the appended claims means any combination of one or more of the associated listed items and all possible combinations, and includes such combinations.

[0139] As used in this application specification and the appended claims, the term "if" may be interpreted, depending on the context, as "when," "once," "in response to determination," or "in response to detection." Similarly, the phrase "if determined" or "if detected [the described condition or event]" may be interpreted, depending on the context, as meaning "once determined," "in response to determination," "once detected [the described condition or event]," or "in response to detection [the described condition or event]."

[0140] Furthermore, in the description of this application and the appended claims, the terms "first," "second," "third," etc., are used only for distinguishing descriptions and should not be construed as indicating or implying relative importance. It should also be understood that although the terms "first," "second," etc., are used in the text to describe various elements in some embodiments of this application, these elements should not be limited by these terms. These terms are merely used to distinguish one element from another. For example, a first table may be named a second table, and similarly, a second table may be named a first table, without departing from the scope of the various described embodiments. Both the first table and the second table are tables, but they are not the same table.

[0141] References to "one embodiment" or "some embodiments" as described in this specification mean that one or more embodiments of this application include a specific feature, structure, or characteristic described in connection with that embodiment. Therefore, the phrases "in one embodiment," "in some embodiments," "in other embodiments," "in still other embodiments," etc., appearing in different parts of this specification do not necessarily refer to the same embodiment, but rather mean "one or more, but not all, embodiments," unless otherwise specifically emphasized. The terms "comprising," "including," "having," and variations thereof mean "including but not limited to," unless otherwise specifically emphasized.

[0142] The motion trajectory planning method provided in this application can be applied to terminal devices such as mobile phones, tablets, wearable devices, in-vehicle devices, augmented reality (AR) / virtual reality (VR) devices, laptops, ultra-mobile personal computers (UMPCs), netbooks, and personal digital assistants (PDAs). This application does not impose any restrictions on the specific type of terminal device.

[0143] For example, the terminal device may be a station (STAION, ST) in a WLAN, a cellular phone, a cordless phone, a Session Initiation Protocol (SIP) phone, a Wireless Local Loop (WLL) station, a Personal Digital Assistant (PDA) device, a handheld device with wireless communication capabilities, a computing device or other processing device connected to a wireless modem, an in-vehicle device, a vehicle networking terminal, a computer, a laptop computer, a handheld communication device, a handheld computing device, a satellite wireless device, a wireless modem card, a set-top box (STB), customer premises equipment (CPE), and / or other devices used for communication over a wireless system, as well as next-generation communication systems, such as mobile terminals in 5G networks or mobile terminals in future evolved Public Land Mobile Network (PLMN) networks.

[0144] As an example and not a limitation, when the terminal device is a wearable device, the term "wearable device" can also refer to any device that utilizes wearable technology to intelligently design and develop everyday wearables, such as glasses, gloves, watches, clothing, and shoes. Wearable devices are portable devices worn directly on the body or integrated into a user's clothing or accessories. Wearable devices are not merely hardware devices; they achieve powerful functions through software support, data interaction, and cloud interaction. Broadly defined, wearable smart devices include those with comprehensive functions, large sizes, and the ability to perform complete or partial functions without relying on a smartphone, such as smartwatches or smart glasses, as well as those focused on a specific application function that require interaction with other devices such as smartphones, such as various smart bracelets and smart jewelry for vital sign monitoring.

[0145] Figure 9 This is a schematic diagram of the structure of a terminal device provided in an embodiment of this application. For example... Figure 9 As shown, the terminal device 9 of this embodiment includes: at least one processor 90 ( Figure 9 (Only one is shown in the image) a memory 91, which stores a computer program 92 that can run on the processor 90. When the processor 90 executes the computer program 92, it implements the steps in the various motion trajectory planning method embodiments described above, for example... Figure 1 Steps S101 to S105 are shown. Alternatively, when the processor 90 executes the computer program 92, it implements the functions of each module / unit in the above-described device embodiments, for example... Figure 8 The functions of modules 810 to 850 are shown.

[0146] The terminal device 9 can be a desktop computer, laptop, handheld computer, or cloud server, etc. The terminal device may include, but is not limited to, a processor 90 and a memory 91. Those skilled in the art will understand that... Figure 9 This is merely an example of terminal device 9 and does not constitute a limitation on terminal device 9. It may include more or fewer components than shown, or combine certain components, or different components. For example, the terminal device may also include input transmission devices, network access devices, buses, etc.

[0147] The processor 90 may be a Central Processing Unit (CPU), or other general-purpose processors, digital signal processors (DSPs), application-specific integrated circuits (ASICs), field-programmable gate arrays (FPGAs), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. A general-purpose processor may be a microprocessor or any conventional processor.

[0148] In some embodiments, the memory 91 may be an internal storage unit of the terminal device 9, such as a hard disk or memory of the terminal device 9. The memory 91 may also be an external storage device of the terminal device 9, such as a plug-in hard disk, smart media card (SMC), secure digital card (SD), flash card, etc., equipped on the terminal device 9. Furthermore, the memory 91 may include both internal and external storage units of the terminal device 9. The memory 91 is used to store the operating system, applications, bootloader, data, and other programs, such as the program code of the computer program. The memory 91 can also be used to temporarily store data that has been sent or will be sent.

[0149] Furthermore, the functional units in the various embodiments of this application can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit. The integrated unit can be implemented in hardware or as a software functional unit.

[0150] This application also provides a terminal device, which includes at least one memory, at least one processor, and a computer program stored in the at least one memory and executable on the at least one processor. When the processor executes the computer program, it causes the terminal device to implement the steps in any of the above method embodiments.

[0151] This application also provides a computer-readable storage medium storing a computer program that, when executed by a processor, implements the steps described in the various method embodiments above.

[0152] This application provides a computer program product that, when run on a terminal device, enables the terminal device to implement the steps described in the various method embodiments above.

[0153] If the integrated module / unit is implemented as a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, all or part of the processes in the methods of the above embodiments can also be implemented by a computer program instructing related hardware. The computer program can be stored in a computer-readable storage medium, and when executed by a processor, it can implement the steps of the various method embodiments described above. The computer program includes computer program code, which can be in the form of source code, object code, executable files, or certain intermediate forms. The computer-readable medium can include: any entity or device capable of carrying the computer program code, a recording medium, a USB flash drive, a portable hard drive, a magnetic disk, an optical disk, a computer memory, a read-only memory (ROM), a random access memory (RAM), an electrical carrier signal, a telecommunication signal, and a software distribution medium, etc.

[0154] In the above embodiments, the descriptions of each embodiment have different focuses. For parts that are not described in detail or recorded in a certain embodiment, please refer to the relevant descriptions of other embodiments.

[0155] Those skilled in the art will recognize that the units and algorithm steps of the various examples described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are implemented in hardware or software depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementation should not be considered beyond the scope of this application.

[0156] The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the units can be selected to achieve the purpose of this embodiment according to actual needs.

[0157] The above-described embodiments are only used to illustrate the technical solutions of this application, and are not intended to limit them. Although this application has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features. Such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of this application, and should all be included within the protection scope of this application.

Claims

1. A motion trajectory planning method, characterized in that, include: Obtain initial motion trajectory planning information; Based on the initial motion trajectory planning information and the preset safety zone information, intermediate motion trajectory planning information is generated; Based on the preset intermediate motion trajectory optimization model, the intermediate motion trajectory planning information is optimized to generate multiple intermediate motion trajectory optimization information. Based on the multiple intermediate motion trajectory optimization information, the preset safety area information, and the preset motion trajectory prediction model, multiple candidate motion trajectory planning information are obtained. Based on a preset motion trajectory filtering model, the multiple candidate motion trajectory planning information is filtered to generate target motion trajectory planning information. The step of generating intermediate motion trajectory planning information based on the initial motion trajectory planning information and the preset safety area information specifically includes: Based on the initial motion trajectory planning information, time information and position coordinate information are used to generate the initial motion trajectory temporal coordinate information; Based on the preset safe zone information, time information and location coordinate information are used to generate safe zone temporal coordinate information; Based on the initial motion trajectory time coordinate information and the safe area time coordinate information, generate safe area position coordinate occupancy information; Based on the location coordinates of the safe zone and the preset safe zone information, the initial motion trajectory planning information is smoothed to generate intermediate motion trajectory planning information. The step of optimizing the intermediate motion trajectory planning information according to the preset intermediate motion trajectory optimization model to generate multiple intermediate motion trajectory optimization information specifically includes: Based on the intermediate motion trajectory planning information, coordinate position information is extracted to generate multiple intermediate motion trajectory coordinate position information and multiple intermediate motion trajectory line segment information; Based on the coordinate position information of the multiple intermediate motion trajectories and the information of the multiple intermediate motion trajectory line segments, the intermediate motion curvature angle information is calculated; When the intermediate motion curvature angle information is less than the preset intermediate motion curvature angle threshold, the coordinate position information of the multiple intermediate motion trajectories is optimized according to the preset intermediate motion trajectory optimization model to generate multiple optimized coordinate position information of intermediate motion trajectories. Based on the optimized coordinate position information of the multiple intermediate motion trajectories, multiple intermediate motion trajectory optimization information is generated; The step of obtaining multiple candidate motion trajectory planning information based on the multiple intermediate motion trajectory optimization information, the preset safety area information, and the preset motion trajectory prediction model specifically includes: Obtain spatial coordinate information for predicted motion trajectory; Based on the position coordinate information of the multiple intermediate motion trajectory optimization information and the spatial position coordinate information of the motion trajectory prediction, multiple motion trajectory prediction spatial position coordinate occupancy status information are generated. Based on the spatial location coordinate occupancy status information of the multiple motion trajectory predictions and the preset motion trajectory prediction model, multiple candidate motion trajectory planning information are obtained.

2. The motion trajectory planning method as described in claim 1, characterized in that, The step of filtering the multiple candidate motion trajectory planning information based on the preset motion trajectory filtering model to generate target motion trajectory planning information specifically includes: Based on the multiple candidate motion trajectory planning information, the preset safe area indication function, and the preset motion trajectory safety penalty coefficient, the motion trajectory safety characterization information is calculated. Based on a preset motion trajectory screening model, the multiple candidate motion trajectory planning information are screened according to the motion trajectory quality characterization information to generate target motion trajectory planning information.

3. The motion trajectory planning method as described in claim 2, characterized in that, The step of calculating the motion trajectory safety representation information based on the multiple candidate motion trajectory planning information, the preset safe area indication function, and the preset motion trajectory safety penalty coefficient specifically includes: Based on the multiple candidate motion trajectory planning information, the preset motion trajectory safety penalty coefficient, the preset motion trajectory smoothness penalty coefficient, and the preset motion trajectory deviation coefficient, motion trajectory quality characterization information is calculated. The coordinate position information of the multiple candidate motion trajectory planning information is extracted to generate multiple candidate motion trajectory coordinate position information and multiple candidate motion trajectory line segment information; Based on the coordinate position information and line segment information of the multiple candidate motion trajectories, the curvature angle information of the multiple candidate motion trajectories is calculated. Based on the curvature angle information of the multiple candidate motion trajectories and the preset motion trajectory smoothness penalty coefficient, motion trajectory smoothness characterization information is calculated; The coordinate position information of the initial motion trajectory planning information is extracted to generate multiple initial motion trajectory coordinate position information. Calculate the distance between the coordinate position information of the multiple candidate motion trajectories and the coordinate position information of the multiple initial motion trajectories to obtain the motion trajectory coordinate position offset information; Based on the motion trajectory coordinate position offset information and the preset motion trajectory offset degree coefficient, the motion trajectory offset degree characterization information is calculated; Based on the motion trajectory safety characterization information, motion trajectory smoothness characterization information, and motion trajectory deviation characterization information, motion trajectory quality characterization information is calculated.

4. The motion trajectory planning method as described in claim 2, characterized in that, The step of filtering multiple candidate motion trajectory planning information based on the preset motion trajectory screening model and generating target motion trajectory planning information according to motion trajectory quality characterization information specifically includes: Based on the preset motion trajectory screening model, the preset motion trajectory screening model is optimized to obtain the optimized motion trajectory screening model. Based on the optimized motion trajectory screening model, the multiple candidate motion trajectory planning information are screened according to the motion trajectory quality characterization information to generate target motion trajectory planning information.

5. A motion trajectory planning device, characterized in that, include: The initial motion trajectory planning information acquisition module is used to acquire initial motion trajectory planning information; The intermediate motion trajectory planning information generation module is used to generate intermediate motion trajectory planning information based on the initial motion trajectory planning information and the preset safety area information. The intermediate motion trajectory optimization information generation module is used to optimize the intermediate motion trajectory planning information according to the preset intermediate motion trajectory optimization model, and generate multiple intermediate motion trajectory optimization information. The candidate motion trajectory planning information generation module is used to obtain multiple candidate motion trajectory planning information based on the multiple intermediate motion trajectory optimization information, the preset safety area information, and the preset motion trajectory prediction model. The target motion trajectory planning information generation module is used to filter the multiple candidate motion trajectory planning information based on a preset motion trajectory filtering model and generate target motion trajectory planning information. The step of generating intermediate motion trajectory planning information based on the initial motion trajectory planning information and the preset safety area information specifically includes: Based on the initial motion trajectory planning information, time information and position coordinate information are used to generate the initial motion trajectory temporal coordinate information; Based on the preset safe zone information, time information and location coordinate information are used to generate safe zone temporal coordinate information; Based on the initial motion trajectory time coordinate information and the safe area time coordinate information, generate safe area position coordinate occupancy information; Based on the location coordinates of the safe zone and the preset safe zone information, the initial motion trajectory planning information is smoothed to generate intermediate motion trajectory planning information. The step of optimizing the intermediate motion trajectory planning information according to the preset intermediate motion trajectory optimization model to generate multiple intermediate motion trajectory optimization information specifically includes: Based on the intermediate motion trajectory planning information, coordinate position information is extracted to generate multiple intermediate motion trajectory coordinate position information and multiple intermediate motion trajectory line segment information; Based on the coordinate position information of the multiple intermediate motion trajectories and the information of the multiple intermediate motion trajectory line segments, the intermediate motion curvature angle information is calculated; When the intermediate motion curvature angle information is less than the preset intermediate motion curvature angle threshold, the coordinate position information of the multiple intermediate motion trajectories is optimized according to the preset intermediate motion trajectory optimization model to generate multiple optimized coordinate position information of intermediate motion trajectories. Based on the optimized coordinate position information of the multiple intermediate motion trajectories, multiple intermediate motion trajectory optimization information is generated; The step of obtaining multiple candidate motion trajectory planning information based on the multiple intermediate motion trajectory optimization information, the preset safety area information, and the preset motion trajectory prediction model specifically includes: Obtain spatial coordinate information for predicted motion trajectory; Based on the position coordinate information of the multiple intermediate motion trajectory optimization information and the spatial position coordinate information of the motion trajectory prediction, multiple motion trajectory prediction spatial position coordinate occupancy status information are generated. Based on the spatial location coordinate occupancy status information of the multiple motion trajectory predictions and the preset motion trajectory prediction model, multiple candidate motion trajectory planning information are obtained.

6. A terminal device, characterized in that, The terminal device includes a memory and a processor. The memory stores a computer program that can run on the processor. When the processor executes the computer program, it implements the steps of the method as described in any one of claims 1 to 4.

7. A computer-readable storage medium storing a computer program, characterized in that, When the computer program is executed by a processor, it implements the steps of the method as described in any one of claims 1 to 4.

Images