Path planning and control method and system for autonomous driving vehicles in container ports

Abstract

The present application relates to the field of path planning technology, and discloses a path planning and control method and system for autonomous container port vehicles. The method comprises: collecting environmental data through multi-source sensors and fusing them through spatiotemporal registration to obtain a digital twin model of the port area; optimizing and generating primary and backup path plans based on the model using a three-layer progressive planner; predicting dynamic target trajectories using a temporal attention network to obtain situation assessments; processing assessment data using a distributed predictive control algorithm to generate control instructions; and forming a safety strategy through adaptive safety constraint monitoring of motion states and continuously optimizing it using deep reinforcement learning. The present application enables the safe and reliable operation of autonomous vehicles in complex and changing port environments, and is particularly capable of making accurate path planning and control decisions when faced with dynamic obstacles.

Description

Technical Field

[0001] The present application relates to the field of path planning technology, and in particular to a path planning and control method and system for an autonomous driving vehicle in a container port. Background Art

[0002] Currently, autonomous vehicle navigation and control systems in container ports primarily employ fixed path planning and simple obstacle avoidance mechanisms. These systems typically rely on pre-set waypoints and basic sensor data, employing a single path planning algorithm for navigation. Furthermore, existing autonomous port systems often utilize independent perception, planning, and control modules, with limited information exchange between these modules and relatively fixed control strategies, lacking the ability to adapt to complex and dynamic environments.

[0003] However, this traditional technical solution has obvious shortcomings: first, fixed path planning cannot effectively cope with the frequently changing dynamic obstacles in the port area; second, the simple modular design makes it impossible for the system to fully utilize multi-source sensor data for collaborative decision-making; third, the lack of adaptive safety control mechanisms can easily lead to safety hazards in complex scenarios; finally, existing systems generally lack the ability to learn and optimize historical data, making it difficult to achieve continuous improvement of control strategies. Summary of the Invention

[0004] The present application provides a path planning and control method and system for autonomous container port vehicles, which are used to achieve safe and reliable operation of autonomous container vehicles in complex and changing port environments, especially when facing dynamic obstacles, and can make accurate path planning and control decisions.

[0005] In the first aspect, the present application provides a path planning and control method for an autonomous driving vehicle in a container port, which includes: collecting port area environmental data through a multi-source sensor network, fusing the environmental data through a spatiotemporal registration algorithm, and obtaining a digital twin environment model of the port area; based on the digital twin environment model of the port area, hierarchically optimizing the vehicle path through a three-layer progressive path planner to obtain a path planning scheme including a main path and an alternative path; based on the path planning scheme, using a temporal attention network to perform trajectory prediction and analysis on dynamic targets in the port area to obtain real-time situation assessment data; based on the real-time situation assessment data, using a distributed model predictive control algorithm to perform collaborative control processing on the vehicle to generate a vehicle control instruction sequence; based on the vehicle control instruction sequence, using an adaptive safety constraint module to monitor the vehicle's motion state in real time to form a multi-level safety response strategy, and using a deep reinforcement learning network to continuously optimize the control strategy.

[0006] In a second aspect, the present application provides a path planning and control system for an autonomous container port vehicle, the path planning and control system for the autonomous container port vehicle comprising:

[0007] A fusion module is used to collect port area environmental data through a multi-source sensor network, and fuse the environmental data through a spatiotemporal registration algorithm to obtain a digital twin environment model of the port area;

[0008] A hierarchical module is used to perform hierarchical optimization processing on the vehicle path through a three-layer progressive path planner according to the port area digital twin environment model to obtain a path planning solution including a main path and alternative paths;

[0009] A prediction module is used to perform trajectory prediction analysis on dynamic targets in the port area based on the path planning scheme using a temporal attention network to obtain real-time situation assessment data;

[0010] A control module, configured to perform coordinated control processing on the vehicle based on the real-time situation assessment data using a distributed model predictive control algorithm to generate a vehicle control instruction sequence;

[0011] The monitoring module is used to monitor the vehicle's motion state in real time through the adaptive safety constraint module according to the vehicle control instruction sequence, form a multi-level safety response strategy, and continuously optimize the control strategy using a deep reinforcement learning network.

[0012] A third aspect of the present invention provides a computer device comprising: a memory and at least one processor, wherein the memory stores instructions; the at least one processor calls the instructions in the memory so that the computer device executes the above-mentioned path planning and control method for autonomous driving vehicles in container ports.

[0013] A fourth aspect of the present invention provides a computer-readable storage medium, which stores instructions that, when executed on a computer, enable the computer to execute the above-mentioned path planning and control method for an autonomous container port vehicle.

[0014] In the technical solution provided by this application, port area environmental data is collected through a multi-source sensor network and fused using a spatiotemporal registration algorithm to construct a high-precision digital twin environment model, effectively solving the problem of incomplete and inaccurate data from traditional single sensors and significantly improving the comprehensiveness and accuracy of environmental perception. A three-layer progressive path planner is used to perform hierarchical optimization of vehicle paths, which not only generates the optimal main path but also includes multiple alternative paths, enhancing the system's ability to respond to emergencies and improving the robustness of path planning. The introduction of a temporal attention network enables the system to accurately predict the trajectories of dynamic targets within the port area. Through deep learning and temporal correlation analysis of target motion characteristics, the system's prediction accuracy of dynamic environmental changes is greatly improved. The application of a distributed model predictive control algorithm realizes collaborative control among multiple vehicles and can dynamically adjust the control strategy based on real-time situation assessment data, significantly improving the efficiency and safety of group control. The adaptive safety constraint module cooperates with the continuous optimization mechanism of the deep reinforcement learning network to enable the system to continuously improve the control strategy based on actual operation data, establishing an adaptive evolutionary safety control system. In terms of algorithmic features, the temporal attention network extracts key temporal features of dynamic targets through attention mechanism processing of time series data, making trajectory prediction more accurate; the deep reinforcement learning network continuously improves the intelligence and adaptability of control decisions through continuous interaction with the environment and strategy optimization. The comprehensive application of these algorithmic features significantly enhances the system's adaptability and control performance in complex port environments. Overall, the solution, through the organic combination of multiple innovative algorithms, has built an intelligent and adaptive port autonomous driving control system, achieving all-round optimization of environmental perception, path planning, trajectory prediction, collaborative control, and safety monitoring, significantly improving the operating efficiency and safety of port autonomous driving vehicles. BRIEF DESCRIPTION OF THE DRAWINGS

[0015] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the following briefly introduces the drawings required for use in the description of the embodiments. Obviously, the drawings described below are some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work.

[0016] Figure 1 This is a schematic diagram of an embodiment of a method for path planning and control of an autonomous driving vehicle in a container port according to an embodiment of the present application;

[0017] Figure 2 This is a schematic diagram of an embodiment of a path planning and control system for an autonomous driving vehicle in a container port according to an embodiment of the present application;

[0018] Figure 3It is a schematic block diagram of the structure of a computer device in an embodiment of the present invention. DETAILED DESCRIPTION

[0019] The embodiments of the present application provide a path planning and control method and system for autonomous driving vehicles in container ports. The terms "first", "second", "third", "fourth", etc. (if any) in the specification and claims of this application and the above-mentioned drawings are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that the data used in this way can be interchanged where appropriate so that the embodiments described herein can be implemented in an order other than that illustrated or described herein. In addition, the terms "including" or "having" and any variations thereof are intended to cover non-exclusive inclusions. For example, a process, method, system, product or device that includes a series of steps or units is not necessarily limited to those steps or units that are clearly listed, but may include other steps or units that are not clearly listed or that are inherent to these processes, methods, products or devices.

[0020] For ease of understanding, the specific process of the embodiment of the present application is described below. Figure 1 In one embodiment of the present application, a method for path planning and control of an autonomous container port vehicle includes:

[0021] Step S101: Collect port area environmental data through a multi-source sensor network, fuse the environmental data through a spatiotemporal registration algorithm, and obtain a digital twin environment model of the port area;

[0022] Step S102: Based on the port area digital twin environment model, the vehicle path is hierarchically optimized using a three-layer progressive path planner to obtain a path planning solution including a primary path and alternative paths;

[0023] Step S103: Based on the path planning scheme, a temporal attention network is used to perform trajectory prediction analysis on dynamic targets in the port area to obtain real-time situation assessment data;

[0024] Step S104: Based on the real-time situation assessment data, the vehicle is cooperatively controlled using a distributed model predictive control algorithm to generate a vehicle control instruction sequence;

[0025] Step S105: Based on the vehicle control command sequence, the vehicle's motion state is monitored in real time via the adaptive safety constraint module to form a multi-level safety response strategy, and the control strategy is continuously optimized using a deep reinforcement learning network.

[0026] It is understood that the execution entity of this application can be the path planning and control system of the container port autonomous driving vehicle, or it can be a terminal or server, and the specific implementation is not limited here. The embodiment of this application is described by taking the server as the execution entity as an example.

[0027] Specifically, port environmental data is collected through a multi-source sensor network. The multi-source sensor network includes a lidar array, a high-definition visual sensor array, and a millimeter-wave radar network, each responsible for collecting different types of environmental information. The lidar array collects three-dimensional point cloud data of the port area, recording the spatial position information of static objects such as containers and fixed facilities; the high-definition visual sensor array obtains real-time image data of the port area, which is used to identify and classify various targets within the port area; and the millimeter-wave radar network detects the speed and direction of moving targets. A spatiotemporal registration algorithm fuses these heterogeneous data, involving three key steps: time synchronization, aligning the data collected by different sensors according to a unified timestamp; spatial registration, converting data from different coordinate systems into a unified reference coordinate system; and data fusion, using a Kalman filter to weight and synthesize the multi-source data to obtain a high-precision digital twin environment model of the port area.

[0028] After obtaining a digital twin model of the port area, a three-layer progressive path planner performs hierarchical optimization of vehicle paths. The first layer is the global path planning layer. Based on the static map of the port area, it uses an improved A* algorithm to generate a coarse-grained path plan. The second layer is the local path planning layer. Incorporating real-time dynamic information, it uses the spatiotemporal RRT* algorithm to locally optimize the path, taking into account the impact of dynamic obstacles. The third layer is the trajectory generation layer, which uses model predictive control to generate smooth driving trajectories. The planning results include a primary path and multiple alternative paths, forming a path planning solution. Based on the path planning solution, a temporal attention network performs trajectory prediction analysis for dynamic targets in the port area. This network processes historical trajectory data using an encoder-decoder structure, with an attention mechanism focusing on key factors influencing target motion. The network inputs include state information such as the target's historical position and velocity, and outputs a predicted future trajectory distribution. A multi-target trajectory association algorithm integrates the motion prediction results of different targets to form real-time situation assessment data.

[0029] The distributed model predictive control algorithm performs collaborative vehicle control based on real-time situation assessment data. The algorithm establishes a vehicle dynamics model within the prediction time domain, transforming multi-vehicle collaborative constraints and obstacle avoidance requirements into optimization problems. Through a distributed computing architecture, multiple vehicles solve their respective control problems in parallel, while also considering the interactions between vehicles. The optimization results generate a sequence of vehicle control instructions, including specific control variables such as speed and steering. The adaptive safety constraint module monitors the vehicle's motion state in real time and dynamically adjusts the safety constraint boundaries based on the vehicle control instruction sequence and current state. This module constructs a multi-level safety response strategy, including normal operation, warning, deceleration, and emergency braking. Through continuous interaction with the environment, the deep reinforcement learning network learns the optimal control strategy and continuously optimizes safety response decisions.

[0030] When an autonomous vehicle needs to transport a container from port area A to area B, the system collects environmental data through multiple sensors. LiDAR detects the position of surrounding static obstacles, visual sensors identify moving forklifts and pedestrians, and millimeter-wave radar measures their speed. A spatiotemporal registration algorithm fuses this data into a unified representation of the environment, and a three-layer progressive path planner generates a primary path and two alternative paths. During driving, a temporal attention network predicts that the forklift ahead may make a turn, and the distributed model predictive control algorithm adjusts the vehicle speed and path accordingly. When potential risks are detected, the adaptive safety constraint module initiates the appropriate level of safety response to ensure the safety and efficiency of the transportation process. Data processing and decision-making throughout the entire process are carried out in real time, and the various modules of the system maintain close data interaction and logical connections.

[0031] In an embodiment of the present application, port area environmental data is collected through a multi-source sensor network, and a spatiotemporal registration algorithm is used for fusion processing to construct a high-precision digital twin environment model, which effectively solves the problem of incomplete and inaccurate data from traditional single sensors and significantly improves the comprehensiveness and accuracy of environmental perception. A three-layer progressive path planner is used to perform hierarchical optimization of vehicle paths, which not only generates the optimal main path, but also includes multiple alternative paths, enhancing the system's ability to respond to emergencies and improving the robustness of path planning. The introduction of the temporal attention network enables the system to accurately predict the trajectory of dynamic targets in the port area. Through deep learning and temporal correlation analysis of target motion characteristics, the system's prediction accuracy of dynamic environmental changes is greatly improved. The application of distributed model predictive control algorithms realizes collaborative control among multiple vehicles, and can dynamically adjust control strategies based on real-time situation assessment data, significantly improving the efficiency and safety of group control. The adaptive safety constraint module cooperates with the continuous optimization mechanism of the deep reinforcement learning network, enabling the system to continuously improve the control strategy based on actual operation data, and establish an adaptive evolutionary safety control system. In terms of algorithmic features, the temporal attention network extracts key temporal features of dynamic targets through attention mechanism processing of time series data, making trajectory prediction more accurate; the deep reinforcement learning network continuously improves the intelligence and adaptability of control decisions through continuous interaction with the environment and strategy optimization. The comprehensive application of these algorithmic features significantly enhances the system's adaptability and control performance in complex port environments. Overall, the solution, through the organic combination of multiple innovative algorithms, has built an intelligent and adaptive port autonomous driving control system, achieving all-round optimization of environmental perception, path planning, trajectory prediction, collaborative control, and safety monitoring, significantly improving the operating efficiency and safety of port autonomous driving vehicles.

[0032] In a specific embodiment, the process of executing step S101 may specifically include the following steps:

[0033] (1) In the step of collecting port area environmental data through a multi-source sensor network, the environmental data is divided into static environmental data stream and dynamic environmental data stream;

[0034] (2) Perform spatial coordinate analysis on the static environmental data stream and establish an index table of port infrastructure locations;

[0035] (3) Perform target identification and classification on dynamic environment data streams to generate real-time target status records;

[0036] (4) Pair the port infrastructure location index table with the real-time target status records by timestamp to form a time series environment feature matrix;

[0037] (5) Perform data cleaning and outlier processing on the time series environmental feature matrix to construct a standardized environmental data set;

[0038] (6) The standardized environmental data set is input into the spatiotemporal registration algorithm, and a digital twin environment model of the port area is generated through feature mapping and data alignment.

[0039] Specifically, the environmental data collected by the multi-source sensor network needs to be processed through data stream classification. According to the data characteristics, the environmental data is divided into static environmental data stream and dynamic environmental data stream. The static environmental data stream contains information that does not change with time, such as the spatial position, size, type, etc. of the fixed facilities of the port, while the dynamic environmental data stream contains target information that changes in real time, such as mobile vehicles and operating equipment. The processing of the static environmental data stream is carried out in the coordinate system 1, and the spatial data collected by different sensors are converted into a unified world coordinate system. The converted data is used to identify the boundaries of different facilities through a spatial clustering algorithm, and the geometric features of the facilities are extracted, including key parameters such as the center point coordinates, orientation angle, and occupied area. Based on these features, an index table of the location of port infrastructure is established. The index table adopts a multi-level tree structure, and each node records the location information, attribute labels, and spatial relationships of the facilities.

[0040] The processing of dynamic environment data streams involves target detection and tracking. Target recognition and classification are performed on sensor data, and the detected dynamic targets are divided into different categories such as autonomous vehicles, manually operated vehicles, and operating machinery. Motion state parameters such as position, speed, and acceleration are extracted for each target, and the target's type label, size information, and behavioral characteristics are recorded. This information constitutes a real-time target status record, and each record has an accurate timestamp. The time series matching process associates the port infrastructure location index table with the real-time target status record. The data is aligned according to the timestamp to establish a time series environment feature matrix. Each row of the matrix represents a time point, and the columns contain static environment information and dynamic target status to form a scene description. There are spatial and temporal correlations between the matrix elements, reflecting the characteristics of the port environment evolving over time.

[0041] Data cleaning is performed on the time series environmental feature matrix, mainly dealing with outliers and missing values. Outlier detection uses a statistical method to calculate the distribution characteristics of each feature and identify data points that significantly deviate from the normal range. For detected outliers, corrections or elimination are made based on the data of the previous and next moments. Missing values are supplemented by spatiotemporal interpolation methods to maintain data continuity and consistency. The cleaned data is standardized, and features of different dimensions are converted to a unified scale space to construct a standardized environmental data set. The standardized environmental data set is input into the spatiotemporal registration algorithm. The algorithm performs feature mapping, projects environmental features into a high-dimensional feature space, and extracts key features that characterize the environmental state. Data alignment is then performed to establish correspondences between different times and different areas to eliminate spatial and temporal inconsistencies. Through these processes, a digital twin environmental model of the port area is generated, which accurately reflects the real-time status of the port area.

[0042] For example, a lidar scan acquires point cloud data of container stacks within a storage yard, a visual camera captures images of lane markings and traffic signs, and a millimeter-wave radar tracks forklifts operating within the yard. After this raw data is classified, static data includes container location coordinates, stack height, aisle width, and other information, while dynamic data includes the real-time position and speed of the forklifts. A position index table, established through spatial coordinate analysis, records the precise location and occupied space of each container stack, while target recognition and classification generates a record of the forklift's motion trajectory. Time-series matching integrates this information into a feature matrix, where each element has a clear physical meaning. After outlier processing and normalization, it is input into a spatiotemporal registration algorithm. This registration process eliminates errors between different sensors, resulting in a digital twin model that fully describes the static layout and dynamic operating status of the storage yard, providing reliable environmental information for path planning by autonomous vehicles.

[0043] In a specific embodiment, the process of executing step S102 may specifically include the following steps:

[0044] (1) The digital twin environment model of the port area is parsed into a road network topology diagram, the vehicle traffic attributes of the road network nodes are marked, and a basic road network dataset is generated;

[0045] (2) Extract the inter-node distance matrix and turning constraints from the road network basic data set and construct the initial cost function for path planning;

[0046] (3) Assign weights to the initial cost function of path planning according to vehicle traffic priority and generate a multi-level path evaluation standard;

[0047] (4) Decomposing the multi-level path evaluation criteria through a three-layer progressive path planner to form a set of path candidates;

[0048] (5) Perform collision risk assessment and travel time calculation on the candidate path set to select the main path;

[0049] (6) Determine the search range of alternative paths based on the evaluation results of the main path and generate a path planning solution.

[0050] Specifically, the digital twin environment model of the port area needs to be converted into a road network topology diagram. This process uses graphics algorithms to extract road centerlines and identify road network nodes and connection relationships. Each road network node is marked with its traffic attributes, including node type (intersection, turning point, loading and unloading point, etc.), traffic direction, speed limit and other information. This information constitutes the road network basic data set, and the adjacency matrix is used to store the connection relationship between nodes. The node distance matrix is extracted from the road network basic data set to record the shortest path distance between any two nodes. At the same time, turning constraints are extracted, including rules such as prohibited turning and forced turning. Based on these data, the initial cost function of path planning is constructed, taking into account multiple factors such as path length, number of turns, and travel time. The calculation formula of the cost function is as follows:

[0051]

[0052] in, represents the total cost of path p, represents the distance of the i-th path, represents the turning cost of the j-th turning point, represents the speed impact factor of the k-th node, 、 、 is the weight coefficient. The cost function is weighted according to vehicle priority, with higher-priority paths receiving lower cost weights. The multi-level path evaluation criteria include safety, efficiency, and reliability assessments. For each level, corresponding evaluation indicators and thresholds are set to form an evaluation system.

[0053] A three-layer progressive path planner decomposes the evaluation criteria. At the global level, an improved A* algorithm is used to generate the initial path; at the local level, a dynamic spatiotemporal RRT* algorithm is used for path optimization; and at the execution level, trajectory smoothing is used to generate executable motion trajectories. This layered process forms a set of candidate paths, each consisting of a sequence of path points and motion parameters. When evaluating the candidate paths, a collision risk assessment is performed to check whether the path has potential collisions with static obstacles or dynamic targets. Travel time calculations consider the vehicle's dynamic constraints and traffic regulations to determine the expected travel time for each path. A comprehensive evaluation screens the optimal primary path.

[0054] Based on the primary path evaluation results, the search range for alternative paths is determined. A search area is established around the primary path, generating multiple alternative paths that, combined with the primary path, form a path planning solution. The generation of alternative paths fully considers the dynamic changes in road conditions and the emergency response needs in unexpected situations.

[0055] For example, autonomous vehicles need to transport containers from the loading and unloading area of the terminal to the storage area. The port road network is divided into several nodes, including loading and unloading points, intersections, and yard entrances, and the connection relationship between the nodes is established through topological structure analysis. The initial cost of all feasible paths is calculated based on the basic data of the road network, among which there are multiple optional paths from the loading and unloading area to the yard. Considering that it is currently the peak operation period, the main roads are given a higher traffic priority, and the weight coefficient of the cost function is adjusted accordingly. The three-layer progressive planner generates a global path with the shortest distance, and then performs local optimization based on the real-time traffic conditions to generate a smooth motion trajectory. Through collision risk assessment, it is found that a section of the main path is undergoing road maintenance, so alternative paths are searched in the surrounding area to form a complete path planning solution with one main and two backups.

[0056] In a specific embodiment, the process of executing step S103 may specifically include the following steps:

[0057] (1) Extract the spatiotemporal parameter sequence of the path from the path planning scheme and establish a dynamic target observation dataset in the port area;

[0058] (2) Group and label the observation data set according to target type to generate target behavior feature vectors;

[0059] (3) Input the target behavior feature vector into the temporal attention network to extract the target movement regularity representation;

[0060] (4) Construct a traffic flow density distribution map of the port area based on the target movement law representation and mark the high-risk intersection areas;

[0061] (5) Conduct temporal correlation analysis on high-risk intersection areas and generate target conflict warning index;

[0062] (6) Combine the target conflict warning index with the motion law representation to obtain real-time situation assessment data.

[0063] Specifically, the spatiotemporal parameter sequence of the path is extracted from the path planning solution to create a dynamic target observation dataset for the port area. This spatiotemporal parameter sequence is a serialized data structure consisting of the location information and estimated transit time of key points along the vehicle's path. It includes parameters such as latitude and longitude coordinates, timestamps, vehicle speed, and heading angle. These parameters are structured and organized into a standardized data table. This data is then combined with existing dynamic target information in the port area, such as the location and motion status of other vehicles, equipment, and personnel, to create a dynamic target observation dataset for the port area.

[0064] After grouping and labeling the observed dataset by target type, a target behavior feature vector needs to be generated. Target types include forklifts, tractors, gantry cranes, bridge cranes, and operators, each with distinct motion characteristics. By analyzing the historical trajectory data for each target type, characteristic parameters such as speed distribution, acceleration variation, steering preference, and dwell time are extracted to form a multidimensional feature vector. For example, for a forklift target, its feature vector might include numerical representations of multiple dimensions, such as average speed, maximum acceleration, common paths, and operating area.

[0065] Inputting the target behavior feature vector into the temporal attention network (TAN) and extracting a representation of the target's motion patterns is a key step. A TAN is a deep learning network structure specialized for processing time series data. It uses an attention mechanism to assign higher weights to key time points in the input sequence, thereby capturing the temporal dependencies of the target's motion. The network comprises an encoder-decoder structure. The encoder converts the input feature vector into a hidden layer representation. The attention layer calculates the weight distribution at different time steps, and the decoder generates predictions based on the attention weights. This network processing can extract representations of the motion patterns of various targets, including common paths, typical speed curves, and inertial motion patterns.

[0066] Constructing a traffic flow density distribution map for the port area based on target motion patterns and marking high-risk intersection areas is fundamental to achieving safety early warnings. The traffic flow density distribution map is a data visualization structure in the form of a heat map. By gridding the port area, the target density of each grid cell is calculated over different time periods, forming a dynamically changing density field. Spatiotemporal analysis of this density field identifies areas with frequent traffic flow intersections and a high potential for congestion. Risk thresholds are set based on historical accident data and expert experience to mark high-risk intersection areas.

[0067] Conducting temporal correlation analysis of high-risk intersection areas and generating a target conflict warning index is an important means of accident prevention. Temporal correlation analysis assesses potential conflict risk by calculating the probability of simultaneous appearance and temporal overlap of different targets in the same area. The target conflict warning index is a comprehensive risk assessment metric calculated by weighted integration of parameters such as target density, relative speed, predicted trajectory intersection angle, and minimum safe distance. Higher index values indicate greater conflict risk and require more stringent preventative measures. The target conflict warning index is combined with motion pattern representation to generate real-time situation assessment data. During this joint analysis, the conflict warning index is used as the primary risk indicator, combined with uncertainty assessment in the target motion pattern representation and historical abnormal behavior statistics to construct a risk assessment model. This real-time situation assessment data contains multi-dimensional information such as risk level, risk duration, risk change trend, and recommended response measures, providing a basis for subsequent control decisions.

[0068] For example, in an autonomous driving system at a container port, the system extracted a sequence of spatiotemporal parameters for the path of an automated transport vehicle (ATV) from a planned route plan, including the coordinates and estimated transit times of 25 key path points. It also collected position and motion data for 10 forklifts, 8 tractors, and 5 operators in the surrounding area, forming a dynamic target observation dataset. After grouping and labeling the observation dataset, multiple feature vectors were generated. For example, the feature vector for the forklift group included parameters such as an average speed of 3.5 m / s, a maximum acceleration of 0.8 m / s², and a typical working area radius of 32 meters. These feature vectors were input into a temporal attention network, which extracted motion pattern representations for each target. For example, a forklift has a 70% probability of slowing down and maneuvering at the edge of the container yard. Based on these motion pattern representations, the system constructed a traffic flow density distribution map for the port area and identified three high-risk intersections where main roads and branch lines intersect. Through time-series correlation analysis, the system calculated that the target conflict warning index for Intersection Area 2 reached 0.78 (on a scale of 0-1) between 10:30 AM and 11:00 AM, exceeding the high-risk threshold of 0.75. Joint analysis results indicated that this area warranted the initiation of a secondary safety response strategy, including reducing speeds and increasing safe distances for vehicles passing through it, generating real-time situation assessment data.

[0069] In a specific embodiment, the process of executing step S104 may specifically include the following steps:

[0070] (1) Decompose the real-time situation assessment data, extract the vehicle state parameters and motion constraints, and generate the vehicle control parameter set;

[0071] (2) Grouping vehicle control parameter sets according to vehicle task priorities to construct a vehicle collaborative decision space;

[0072] (3) Match the constraints of the control variables in the vehicle collaborative decision space to generate the vehicle collaborative control constraint matrix;

[0073] (4) Optimize the vehicle cooperative control constraint matrix through the distributed model predictive control algorithm to form a vehicle action sequence table;

[0074] (5) Conflict detection and avoidance rule verification based on the vehicle action sequence table, and establishment of a vehicle command priority list;

[0075] (6) Sort and merge the vehicle command priority list in chronological order to generate a vehicle control command sequence.

[0076] Specifically, real-time situation assessment data is decomposed to extract vehicle state parameters and motion constraints, generating a set of vehicle control parameters. Real-time situation assessment data is a multidimensional information structure containing state parameters such as the vehicle's current position, speed, and acceleration, as well as constraints such as the surrounding safety distance and obstacle distribution. The data decomposition process utilizes a feature separation algorithm to classify the raw situation data into three categories based on physical properties: dynamic parameter group, environmental constraint group, and task requirement group. The dynamic parameter group includes speed limits, steering angle limits, and acceleration boundaries; the environmental constraint group includes obstacle avoidance distances, safety margins, and traffic regulations; and the task requirement group includes target location, arrival time, and energy consumption requirements. Through data reorganization and normalization, these parameters are integrated into a standardized set of vehicle control parameters. Grouping the vehicle control parameter sets according to vehicle task priority and constructing a vehicle collaborative decision space is the foundation of multi-vehicle collaboration. Vehicle task priority is a comprehensive assessment of task urgency, importance, and resource utilization, and is typically categorized into four levels: urgent tasks, important tasks, routine tasks, and background tasks. The grouping process utilizes a hierarchical clustering approach, grouping vehicle control parameters with similar priorities into the same decision-making hierarchy, forming a multi-level decision-making structure. The vehicle collaborative decision space is a multidimensional vector space representation, where each dimension represents a decision variable, such as speed selection, path selection, and action timing. All possible decision combinations constitute the decision space. By mapping decision variables of different priorities to corresponding spatial regions, a hierarchical collaborative decision space is established, providing the foundation for subsequent constraint matching.

[0077] Matching constraints on control variables in the vehicle collaborative decision-making space to generate a vehicle collaborative control constraint matrix involves a complex constraint processing process. Control variables include directly manipulable variables such as speed, acceleration, steering angle, and braking force. Constraints include various types, such as physical limitations, safety constraints, and energy consumption constraints. Constraint matching uses a graph-based mapping algorithm to connect each control variable with the relevant constraints to form a constraint network. By analyzing the constraint network, key and redundant constraints are identified, and the constraint structure is optimized. The vehicle collaborative control constraint matrix is a mathematical representation in which each element represents the constraint coefficient of a specific control variable under specific constraints, forming a set of constraint equations.

[0078] The core control step is to optimize the vehicle cooperative control constraint matrix using a distributed model predictive control algorithm to form a vehicle action sequence table. The distributed model predictive control algorithm is an advanced control algorithm that predicts the system behavior over a period of time in the future and optimizes current control decisions based on the predicted results. First, a vehicle dynamics model is established to predict the future state trajectory under different control inputs. Then, an objective function is set, which includes multiple optimization goals such as trajectory tracking error, control input smoothness, and energy consumption minimization. By solving the constrained optimization problem, the optimal control sequence that meets all constraints is obtained. The distributed processing architecture enables multiple vehicle controllers to perform parallel calculations and exchange intermediate results through communication to achieve global optimization. The optimized calculation results form a vehicle action sequence table, which contains time-series action instructions such as specific speed curves, steering angle changes, acceleration and deceleration timing, etc.

[0079] Conflict detection and avoidance rule verification based on the vehicle action sequence table, along with the establishment of a vehicle command priority list, are key elements in ensuring safety. Conflict detection utilizes a spatiotemporal trajectory intersection analysis method, identifying potential conflict points by calculating the minimum distance and temporal overlap between the predicted trajectories of different vehicles. Detected conflicts are resolved using pre-set avoidance rules, such as priority for vehicles on the right, priority for vehicles traveling through, and priority for loaded vehicles. During the avoidance rule verification process, the corresponding rules are applied to each potential conflict, determining which vehicle should yield and which vehicle should take priority, and revising the corresponding action sequence. The vehicle command priority list records the execution priorities and interdependencies of all commands, ensuring that control commands are executed in a reasonable order in the event of a conflict. Temporal sorting utilizes a topological sorting algorithm, determining the final execution order based on inter-command dependencies and execution time points. Command merging combines and simplifies commands with similar timing and operations, reducing the frequency of command switching and improving control smoothness. A vehicle control command sequence is a standardized set of commands sent directly to the vehicle's actuators, containing specific execution times, execution content, and execution parameters.

[0080] For example, in an autonomous driving system at a container port, when multiple automated transport vehicles need to coordinate operations within a container yard, the system first acquires real-time situation assessment data, including the current status and surrounding environment of five vehicles. This data is then decomposed to extract state parameters such as each vehicle's position, speed, and acceleration, as well as constraints such as road width, turning radius, and safety distance, to form a vehicle control parameter set. The five vehicles are then divided into two high-priority vehicles (carrying critical cargo) and three regular-priority vehicles based on mission urgency, creating a collaborative decision space. Within this decision space, constraints are matched against control variables such as speed, path, and scheduling time to generate a collaborative control constraint matrix. A distributed model predictive control algorithm is used to calculate the optimal control strategy, generating a sequence table containing detailed maneuvers for the next 30 seconds. The system then detects a potential merging conflict between two vehicles on a narrow road section and applies a "heavier vehicle first" avoidance rule, adjusting the speed profile of the regular-priority vehicle and delaying its passage, thereby establishing a priority list of instructions. Ultimately, the system sorted and merged these instructions in chronological order, generating a highly optimized vehicle control instruction sequence, enabling all vehicles to complete collaborative tasks efficiently and safely, greatly improving port operation efficiency.

[0081] In a specific embodiment, the process of executing step S105 may specifically include the following steps:

[0082] (1) Construct an instruction execution timing table based on the vehicle control instruction sequence, extract vehicle motion control parameters, and generate vehicle motion state monitoring conditions;

[0083] (2) Input the vehicle motion state monitoring conditions into the adaptive safety constraint module to establish a dynamic safety boundary constraint set;

[0084] (3) Perform boundary detection on the vehicle’s current motion state based on the dynamic safety boundary constraint set and generate safety risk assessment indicators;

[0085] (4) Using security risk assessment indicators as weight coefficients, a hierarchical security response rule base is constructed to form a multi-level security response strategy;

[0086] (5) Pairing the multi-level safety response strategy with historical control data, inputting it into a deep reinforcement learning network, and generating control strategy optimization parameters;

[0087] (6) Dynamically adjust the multi-level safety response strategy through control strategy optimization parameters to complete the vehicle's motion status monitoring.

[0088] Specifically, the timing table contains the execution time, execution order, and execution conditions for each instruction, and also extracts motion control parameters such as speed commands, steering commands, and braking commands. By analyzing the temporal and spatial distribution characteristics of these parameters, motion state monitoring conditions are generated, including specific constraints such as speed thresholds, steering angle ranges, and acceleration limits. After the motion state monitoring conditions are input into the adaptive safety constraint module, a dynamic safety boundary constraint set is established based on the vehicle's dynamic characteristics and road conditions. The constraint set consists of speed safety boundaries, position safety boundaries, and posture safety boundaries. The speed safety boundary defines the vehicle's safe speed range in different scenarios, the position safety boundary determines the minimum safe distance from other vehicles and fixed facilities, and the posture safety boundary limits the vehicle's yaw and roll angles.

[0089] When performing boundary detection on the vehicle's current motion state, the actual motion parameters are compared in real time with the safety boundary constraint set. The deviation between each monitored parameter and the corresponding boundary is calculated, and the severity of each deviation is comprehensively assessed to generate a quantitative safety risk assessment index. This index reflects the safety level of the vehicle's current motion state, with higher values indicating higher risk. Based on the safety risk assessment index, a safety response rule base with multiple response levels is constructed. Different risk indicator ranges correspond to different response strategies, forming a multi-level safety response strategy. The response levels, from low to high, are normal operation, early warning, active intervention, and emergency braking. Each level has clear trigger conditions and response measures.

[0090] Multi-level safety response strategies are paired with historical control data for analysis, extracting metrics for strategy execution effectiveness, including response timeliness, control smoothness, and safety margin. This data serves as training samples and is fed into a deep reinforcement learning network. Through repeated learning and optimization, the network generates optimized control strategy parameters that adapt to different scenarios. The optimized control strategy parameters are then used to dynamically adjust the multi-level safety response strategy. Adjustments include trigger thresholds, response strengths, and execution timing for each level, ensuring that the safety response strategy can be flexibly adjusted based on actual conditions and enabling accurate monitoring of vehicle motion.

[0091] For example, when an autonomous container truck operates in a narrow passage, it establishes a timing table based on the control command sequence, extracting key parameters such as steering angle and driving speed. Taking into account the passage width and road conditions, the adaptive safety constraint module sets a smaller speed limit and a larger safety distance requirement. During driving, real-time boundary detection reveals that the distance between the vehicle and the container stack on one side is approaching the safety boundary, and the safety risk assessment index increases. The multi-level safety response strategy immediately activates the warning level, while the deep reinforcement learning network calculates the optimized steering angle and deceleration strategy based on historical data. The dynamically adjusted response strategy enables the vehicle to smoothly adjust its driving route, always staying within the safety boundary. Data processing and decision adjustments throughout the monitoring process are closely centered on the core goal of safety, and each link maintains a strict logical connection.

[0092] In a specific embodiment, the process of executing step S106 may specifically include the following steps:

[0093] (1) Classify and sort the safety risk assessment indicators according to risk levels and generate a risk classification threshold table;

[0094] (2) Divide the values in the risk grading threshold table into intervals and establish a response trigger condition matrix;

[0095] (3) Construct a security response rule mapping set based on the response trigger condition matrix and mark the rule priority number;

[0096] (4) Associating the rule priority number with the vehicle control constraint condition to form a rule constraint relationship table;

[0097] (5) According to the rule constraint relationship table, the response rules are hierarchically combined to build a security response rule base;

[0098] (6) Prioritize the rule priority numbers according to the security response rule base to form a multi-level security response strategy.

[0099] Specifically, the system begins by categorizing and ranking safety risk assessment indicators, dividing them into different levels based on their degree of danger. These indicators include vehicle motion parameters (such as speed, acceleration, and steering angle), environmental perception data (such as obstacle distance and relative speed), and mission execution status (such as loading and unloading status and battery charge).

[0100] For speed indicators, a baseline speed threshold of 20 km / h is set in the port scenario. When the vehicle speed exceeds this threshold, a risk warning state is entered. The baseline threshold for the acceleration indicator is set at 2 m / s², and the threshold for the steering angle change rate is 30° / s. These kinematic parameters are combined with environmental perception data such as obstacle distance and relative speed to form a multi-dimensional risk assessment system. For example, if the autonomous vehicle detects a dynamic obstacle within 10 meters ahead and the relative speed exceeds 5 km / h, a higher risk level is triggered. Based on these specific indicators, a threshold table containing multiple risk levels is generated. The risk classification threshold table is then divided into intervals to establish a response trigger condition matrix. This interval division adopts a non-uniform division based on risk level, setting different trigger thresholds for different risk indicators. Taking vehicle speed as an example, 20-25 km / h is defined as the first warning interval, 25-30 km / h as the second warning interval, and over 30 km / h as the third warning interval. Obstacle distance is similarly divided into three intervals: 10-15 m, 5-10 m, and 0-5 m. These interval division results are multiplied by the weight coefficients of each indicator to form a response trigger condition matrix.

[0101] Based on the response trigger condition matrix, a safety response rule mapping set is constructed and assigned rule priority numbers. Mapping rules are set based on actual scenario requirements. For example, when a dynamic obstacle is detected within 5 meters ahead, the emergency braking rule is triggered, assigned the highest priority number 1. When the vehicle speed is detected to be about to exceed 25 km / h, the deceleration rule is triggered, assigned priority number 2. This method establishes a clear mapping relationship between various risk scenarios and corresponding control responses. Rule priority numbers are associated with specific vehicle control constraints to form a rule constraint relationship table. Control constraints include parameters such as maximum braking deceleration, maximum steering angular velocity, and minimum safe distance. For example, the emergency braking rule (priority 1) corresponds to a maximum braking deceleration of 3 m / s² and a minimum safe distance of 2 meters; the deceleration rule (priority 2) corresponds to a braking deceleration of 1 m / s². This association transforms abstract rule numbers into specific control parameters. Based on the rule constraint relationship table, response rules are hierarchically combined to construct a safety response rule base. Rule combination considers the coupling effects of multiple risk factors. For example, when both high speed and short distance risk factors exist, more stringent control parameters are selected. The rule base contains three levels: basic security rules, warning rules and emergency rules, and multiple specific response rules are set under each level.

[0102] Based on the security response rule base, the rules are prioritized to form a multi-level security response strategy. Prioritization utilizes a comprehensive assessment based on risk and timeliness to ensure the most appropriate response strategy is triggered promptly in complex scenarios. For example, when an autonomous vehicle encounters an unexpected obstacle in a port channel, the response strategy sequentially activates multiple levels of control commands, including deceleration, avoidance, and braking. Each control command has clear triggering conditions and execution priorities.

[0103] For example, a self-driving container transporter is traveling on a main road in a port area when a forklift, engaged in loading or unloading operations, suddenly appears in front of it. The safety monitoring module detects a distance of 8 meters from the forklift, a relative speed of 10 km / h, and a vehicle speed of 22 km / h. Based on the risk grading threshold table, this scenario triggers both distance (5-10m range) and speed (20-25 km / h range) risk indicators. The response trigger condition matrix combines these two risk factors to determine the need for deceleration and evasive maneuvers. The safety response rule mapping set maps this scenario to an avoidance rule (priority 3) and a deceleration rule (priority 2). The rule constraint relationship table assigns specific control parameters to these two rules: a deceleration of 1.5 m / s² and a steering angular velocity not exceeding 20° / s. The multi-level safety response strategy executes deceleration and evasive maneuvers in order of priority, ensuring the transporter can safely pass through the road section. The entire process embodies a complete logical chain from risk identification to the generation of specific control instructions, fully considering the safe operation requirements of self-driving vehicles in port areas.

[0104] The above describes the path planning and control method of the container port autonomous driving vehicle in the embodiment of the present application. The following describes the path planning and control system of the container port autonomous driving vehicle in the embodiment of the present application. Figure 2 In one embodiment of the present application, a path planning and control system for an autonomous container port vehicle includes:

[0105] The fusion module is used to collect port area environmental data through a multi-source sensor network, fuse the environmental data through a spatiotemporal registration algorithm, and obtain a digital twin environment model of the port area;

[0106] A hierarchical module is used to perform hierarchical optimization of vehicle paths using a three-layer progressive path planner based on the port area digital twin environment model, resulting in a path planning solution that includes a primary path and alternative paths.

[0107] The prediction module is used to perform trajectory prediction analysis on dynamic targets in the port area based on the path planning scheme using the temporal attention network to obtain real-time situation assessment data;

[0108] A control module, configured to perform coordinated control processing on the vehicle based on the real-time situation assessment data using a distributed model predictive control algorithm to generate a vehicle control instruction sequence;

[0109] The monitoring module is used to monitor the vehicle's motion state in real time through the adaptive safety constraint module according to the vehicle control instruction sequence, form a multi-level safety response strategy, and continuously optimize the control strategy using a deep reinforcement learning network.

[0110] Through the collaborative efforts of these components, a multi-source sensor network collects port environmental data and fuses it using a spatiotemporal registration algorithm to construct a high-precision digital twin environment model. This effectively addresses the incomplete and inaccurate data from traditional single sensors, significantly improving the comprehensiveness and accuracy of environmental perception. A three-layer progressive path planner optimizes vehicle paths in a hierarchical manner, generating not only the optimal primary path but also multiple alternative paths. This enhances the system's ability to respond to emergencies and improves the robustness of path planning. The introduction of a temporal attention network enables the system to accurately predict the trajectories of dynamic targets within the port. Through deep learning of target motion characteristics and temporal correlation analysis, the system significantly improves its accuracy in predicting dynamic environmental changes. The application of a distributed model predictive control algorithm enables coordinated control among multiple vehicles, dynamically adjusting control strategies based on real-time situation assessment data, significantly improving the efficiency and safety of swarm control. The adaptive safety constraint module, coupled with the continuous optimization mechanism of a deep reinforcement learning network, enables the system to continuously refine control strategies based on actual operational data, establishing an adaptive and evolving safety control system. In terms of algorithmic features, the temporal attention network extracts key temporal features of dynamic targets through attention mechanism processing of time series data, making trajectory prediction more accurate; the deep reinforcement learning network continuously improves the intelligence and adaptability of control decisions through continuous interaction with the environment and strategy optimization. The comprehensive application of these algorithmic features significantly enhances the system's adaptability and control performance in complex port environments. Overall, the solution, through the organic combination of multiple innovative algorithms, has built an intelligent and adaptive port autonomous driving control system, achieving all-round optimization of environmental perception, path planning, trajectory prediction, collaborative control, and safety monitoring, significantly improving the operating efficiency and safety of port autonomous driving vehicles.

[0111] Reference Figure 3 In an embodiment of the present invention, a computer device is also provided. The computer device may be a server, and its internal structure may be as follows: Figure 3As shown. The computer device includes a processor, memory, display screen, input device, network interface and database connected via a system bus. The processor of the computer design is used to provide computing and control capabilities. The memory of the computer device includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system, a computer program and a database. The internal memory provides an environment for the operation of the operating system and computer program in the non-volatile storage medium. The database of the computer device is used to store the corresponding data in this embodiment. The network interface of the computer device is used to communicate with an external terminal via a network connection. When the computer program is executed by the processor, the above method is implemented.

[0112] Those skilled in the art will understand that Figure 3 The structure shown in the figure is merely a block diagram of a portion of the structure related to the solution of the present invention and does not constitute a limitation on the computer device to which the solution of the present invention is applied.

[0113] An embodiment of the present invention further provides a computer-readable storage medium having a computer program stored thereon, which implements the above-described method when executed by a processor. It is understood that the computer-readable storage medium in this embodiment can be a volatile readable storage medium or a non-volatile readable storage medium.

[0114] Those skilled in the art will appreciate that all or part of the processes in the above-described method embodiments can be implemented by instructing the relevant hardware using a computer program. The computer program can be stored in a non-volatile computer-readable storage medium. When executed, the computer program can include the processes of the above-described method embodiments. Any reference to memory, storage, database, or other media provided herein and used in the embodiments may include non-volatile and / or volatile memory. Non-volatile memory may include read-only memory (ROM), programmable ROM (PROM), electrically programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM), or flash memory. Volatile memory may include random access memory (RAM) or external cache memory. By way of illustration and not limitation, RAM is available in various forms, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), double-speed SDRAM (SSRSDRAM), enhanced SDRAM (ESDRAM), synchronous link DRAM (SLDRAM), RAMbus direct RAM (RDRAM), direct RAMbus dynamic RAM (DRDRAM), and RAMbus dynamic RAM.

[0115] Those skilled in the art will clearly understand that, for the convenience and brevity of description, the specific working processes of the above-described systems, systems and units can refer to the corresponding processes in the aforementioned method embodiments and will not be repeated here.

[0116] If the integrated 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, the technical solution of the present invention, or the portion that contributes to the prior art, or all or part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions for enabling a computer device (which can be a personal computer, server, or network device, etc.) to execute all or part of the steps of the method described in each embodiment of the present invention. The aforementioned storage medium includes various media that can store program code, such as a USB flash drive, a mobile hard drive, a read-only memory (ROM), a random access memory (RAM), a magnetic disk, or an optical disk.

[0117] As described above, the above embodiments are only used to illustrate the technical solutions of the present application, rather than to limit them. Although the present application has been described in detail with reference to the above embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the above embodiments, or make equivalent replacements for some of the technical features therein. However, these modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the embodiments of the present application.

Claims

1. A path planning and control method for an autonomous driving vehicle in a container port, characterized in that: The path planning and control method for the container port autonomous driving vehicle includes: The port area environmental data is collected through a multi-source sensor network, and the environmental data is fused and processed through a spatiotemporal registration algorithm to obtain a digital twin environmental model of the port area; According to the digital twin environment model of the port area, the vehicle path is hierarchically optimized through a three-layer progressive path planner to obtain a path planning scheme including a main path and alternative paths, including: parsing the digital twin environment model of the port area into a road network topology diagram, marking the vehicle traffic attributes of the road network nodes, and generating a road network basic data set; extracting the inter-node distance matrix and steering constraints from the road network basic data set to construct an initial cost function for path planning; weighting the initial cost function for path planning according to the vehicle traffic priority to generate a multi-level path evaluation standard; decomposing the multi-level path evaluation standard through the three-layer progressive path planner to form a path candidate set; performing collision risk assessment and travel time calculation on the path candidate set to screen out the main path; determining the alternative path search range based on the evaluation result of the main path to generate the path planning scheme; Based on the path planning scheme, a temporal attention network is used to perform trajectory prediction analysis on dynamic targets in the port area to obtain real-time situation assessment data, including: extracting a path spatiotemporal parameter sequence from the path planning scheme to establish a dynamic target observation data set in the port area; grouping and labeling the observation data set according to target type to generate a target behavior feature vector; inputting the target behavior feature vector into the temporal attention network to extract a target motion law representation; constructing a port area traffic flow density distribution map based on the target motion law representation and marking high-risk intersection areas; performing temporal correlation analysis on the high-risk intersection areas to generate a target conflict warning index; and jointly analyzing the target conflict warning index with the motion law representation to obtain the real-time situation assessment data; Based on the real-time situation assessment data, the vehicles are cooperatively controlled by a distributed model predictive control algorithm to generate a vehicle control instruction sequence, including: performing data decomposition on the real-time situation assessment data, extracting vehicle state parameters and motion constraints, and generating a vehicle control parameter set; grouping the vehicle control parameter set according to vehicle task priorities to construct a vehicle cooperative decision space; performing constraint matching on the control variables in the vehicle cooperative decision space to generate a vehicle cooperative control constraint matrix; optimizing and calculating the vehicle cooperative control constraint matrix via the distributed model predictive control algorithm to form a vehicle action sequence table; performing conflict detection and avoidance rule verification based on the vehicle action sequence table to establish a vehicle instruction priority list; sorting and merging the vehicle instruction priority list in chronological order to generate the vehicle control instruction sequence; According to the vehicle control instruction sequence, the vehicle's motion state is monitored in real time through the adaptive safety constraint module to form a multi-level safety response strategy, and the control strategy is continuously optimized using a deep reinforcement learning network.

2. The path planning and control method for an autonomous container port vehicle according to claim 1, characterized in that: The port area environmental data is collected through a multi-source sensor network, and the environmental data is fused and processed through a spatiotemporal registration algorithm to obtain a port area digital twin environment model, including: In the step of collecting port area environmental data via the multi-source sensor network, the environmental data is divided into a static environmental data stream and a dynamic environmental data stream; Perform spatial coordinate analysis on the static environment data stream to establish a port infrastructure location index table; perform target identification and classification on the dynamic environment data stream to generate real-time target status records; Pairing the port infrastructure location index table with the real-time target status record according to timestamps to form a time series environment feature matrix; Performing data cleaning and outlier processing on the time series environmental feature matrix to construct a standardized environmental data set; The standardized environmental data set is input into the spatiotemporal registration algorithm, and the digital twin environment model of the port area is generated through feature mapping and data alignment.

3. The path planning and control method for an autonomous container port vehicle according to claim 1, characterized in that: The vehicle control instruction sequence is used to monitor the vehicle's motion state in real time via an adaptive safety constraint module to form a multi-level safety response strategy, and a deep reinforcement learning network is used to continuously optimize the control strategy, including: Building an instruction execution timing table based on the vehicle control instruction sequence, extracting vehicle motion control parameters, and generating vehicle motion state monitoring conditions; Inputting the vehicle motion state monitoring condition into the adaptive safety constraint module to establish a dynamic safety boundary constraint set; Perform boundary detection on the current motion state of the vehicle according to the dynamic safety boundary constraint set to generate a safety risk assessment index; Using the security risk assessment indicators as weight coefficients, a hierarchical security response rule base is constructed to form the multi-level security response strategy; Pairing the multi-level safety response strategy with historical control data, inputting the data into the deep reinforcement learning network, and generating control strategy optimization parameters; The multi-level safety response strategy is dynamically adjusted by optimizing the control strategy parameters to complete the motion state monitoring of the vehicle.

4. The path planning and control method for an autonomous container port vehicle according to claim 3, characterized in that: The method of constructing a hierarchical security response rule base using the security risk assessment indicators as weight coefficients to form the multi-level security response strategy includes: Classify and sort the security risk assessment indicators according to risk levels to generate a risk classification threshold table; Divide the values in the risk classification threshold table into intervals to establish a response trigger condition matrix; construct a security response rule mapping set based on the response trigger condition matrix and mark the rule priority sequence number; Associating the rule priority sequence number with the vehicle control constraint condition to form a rule constraint relationship table; The response rules are hierarchically combined according to the rule constraint relationship table to construct a security response rule library; The rule priority numbers are prioritized according to the security response rule base to form the multi-level security response strategy.

5. A path planning and control system for an autonomous container port vehicle, for implementing the path planning and control method for an autonomous container port vehicle according to any one of claims 1 to 4, characterized in that: The path planning and control system for the container port autonomous driving vehicle includes: A fusion module is used to collect port area environmental data through a multi-source sensor network, and fuse the environmental data through a spatiotemporal registration algorithm to obtain a digital twin environment model of the port area; A hierarchical module is used to perform hierarchical optimization processing on the vehicle path through a three-layer progressive path planner according to the digital twin environment model of the port area to obtain a path planning scheme including a main path and alternative paths, including: parsing the digital twin environment model of the port area into a road network topology structure diagram, marking the vehicle traffic attributes of the road network nodes, and generating a road network basic data set; extracting the node distance matrix and steering constraints from the road network basic data set to construct an initial cost function for path planning; weighting the initial cost function for path planning according to the vehicle traffic priority to generate a multi-level path evaluation standard; decomposing the multi-level path evaluation standard through the three-layer progressive path planner to form a path candidate set; performing collision risk assessment and travel time calculation on the path candidate set to screen out the main path; determining the alternative path search range based on the evaluation result of the main path, and generating the path planning scheme; A prediction module is used to perform trajectory prediction analysis on dynamic targets in the port area based on the path planning scheme using a temporal attention network to obtain real-time situation assessment data, including: extracting a path spatiotemporal parameter sequence from the path planning scheme to establish a dynamic target observation data set in the port area; grouping and labeling the observation data set according to target type to generate a target behavior feature vector; inputting the target behavior feature vector into the temporal attention network to extract a target motion law representation; constructing a port area traffic flow density distribution map based on the target motion law representation and marking high-risk intersection areas; performing temporal correlation analysis on the high-risk intersection areas to generate a target conflict warning index; and jointly analyzing the target conflict warning index with the motion law representation to obtain the real-time situation assessment data; A control module is used to perform collaborative control processing on vehicles based on the real-time situation assessment data through a distributed model predictive control algorithm to generate a vehicle control instruction sequence, including: performing data decomposition on the real-time situation assessment data, extracting vehicle state parameters and motion constraints, and generating a vehicle control parameter set; grouping the vehicle control parameter set according to vehicle task priorities to construct a vehicle collaborative decision space; performing constraint matching on control variables in the vehicle collaborative decision space to generate a vehicle collaborative control constraint matrix; optimizing and calculating the vehicle collaborative control constraint matrix through the distributed model predictive control algorithm to form a vehicle action sequence table; performing conflict detection and avoidance rule verification based on the vehicle action sequence table to establish a vehicle instruction priority list; sorting and merging the vehicle instruction priority list in chronological order to generate the vehicle control instruction sequence; The monitoring module is used to monitor the vehicle's motion state in real time through the adaptive safety constraint module according to the vehicle control instruction sequence, form a multi-level safety response strategy, and continuously optimize the control strategy using a deep reinforcement learning network.

6. A computer device, characterized in that: It includes a memory and a processor, the memory stores a computer program that can be run on the processor, and is characterized in that when the processor executes the computer program, it implements the path planning and control method of the container port autonomous driving vehicle according to any one of claims 1 to 4.

7. A computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, causes the processor to execute the path planning and control method for an autonomous container port vehicle according to any one of claims 1 to 4.

Images