Intelligent logistics scheduling method and device based on dynamic weight and medium

Abstract

The invention discloses an intelligent logistics scheduling method and device based on a dynamic weight and a medium, and the method comprises the steps: carrying out the fusion processing of the operation environment data of a logistics carrier, so as to construct a dynamic environment model, and carrying out the fusion processing of the operation environment data of the logistics carrier based on a reinforcement learning model according to a real-time environment state and a multi-target optimization demand. Dynamically adjusting the weight vector of each logistics parameter in the logistics carrier; predicting the space-time trajectory of each logistics carrier based on the kinematic model of the logistics carrier, mapping the space-time trajectory to a three-dimensional grid, identifying a potential conflict area between the logistics carriers, and calculating a corresponding conflict severity index; based on the dynamic weight vector and the conflict severity index, generating an obstacle avoidance path of the logistics carrier by adopting a global path re-planning algorithm, and optimizing the obstacle avoidance path through a local track to generate a corresponding smooth track; and executing a scheduling instruction corresponding to the smooth trajectory, and collecting an execution feedback result in real time to update the reinforcement learning model parameters and the dynamic weight strategy.

Description

Technical Field

[0001] The present application relates to the field of smart logistics technology, and in particular to a smart logistics scheduling method, equipment and medium based on dynamic weights. Background Art

[0002] As smart logistics systems develop towards automation and high density, multi-AGV collaborative scheduling technology has become key to improving warehouse operational efficiency. In complex and dynamic environments, traditional AGV scheduling methods often use fixed rules or static weights for path planning, which makes it difficult to adapt to real-time changes in traffic density, task priorities, and equipment status. Especially in multi-objective optimization scenarios, the fixed weight allocation mechanism cannot flexibly balance conflicting objectives such as path length and energy efficiency, which can easily lead to local congestion or waste of resources. In addition, existing systems lack the ability to deeply integrate dynamic environmental perception data, and the problems of accumulated positioning errors and noise interference are prominent, further limiting the accuracy of scheduling decisions. Summary of the Invention

[0003] The embodiments of the present application provide a smart logistics scheduling method, device and medium based on dynamic weights to solve the above technical problems.

[0004] On the one hand, the embodiment of the present application provides a smart logistics scheduling method based on dynamic weights, including: The operating environment data of the logistics vehicle is integrated and processed to construct a dynamic environment model. Based on the reinforcement learning model, the weight vector of each logistics parameter in the logistics vehicle is dynamically adjusted according to the real-time environmental status and multi-objective optimization requirements; Predicting the spatiotemporal trajectory of each logistics vehicle based on its kinematic model, mapping the spatiotemporal trajectory to a three-dimensional grid, identifying potential conflict areas between logistics vehicles, and calculating corresponding conflict severity indicators; Based on the dynamic weight vector and the conflict severity index, and using a global path replanning algorithm, an obstacle avoidance path for the logistics vehicle is generated, and the obstacle avoidance path is optimized through local trajectories to generate a corresponding smooth trajectory; Execute the scheduling instructions corresponding to the smooth trajectory and collect execution feedback results in real time to update the reinforcement learning model parameters and dynamic weight strategy.

[0005] In one implementation of the present application, the operation environment data of the logistics vehicle is fused and processed, specifically including: Collecting real-time operating environment data of logistics vehicles through multi-source sensors; the operating environment data includes lidar point cloud data, ultra-wideband base station signals, equipment status parameters and task queue information; De-noising the LiDAR point cloud data based on a voxel filtering algorithm, and matching the historical point cloud map with an iterative closest point algorithm to update the positioning information of the logistics vehicle in real time; fusing the ultra-wideband base station signal with inertial measurement unit data and wheel odometer information in the device state parameters, constructing a three-dimensional network through extended Kalman filtering, and dynamically adjusting the fusion weight according to signal strength; The task queue information is received in real time through a message queue telemetry transmission protocol to parse the task priority and deadline in the task queue information and monitor the battery power and load status of the logistics vehicle.

[0006] In one implementation of the present application, the spatiotemporal trajectory of each logistics vehicle is predicted based on the kinematic model of the logistics vehicle, and the spatiotemporal trajectory is mapped to a three-dimensional grid. Potential conflict areas between logistics vehicles are identified, and corresponding conflict severity indicators are calculated, specifically including: Based on the historical speed, historical acceleration, and path point sequence of the logistics vehicle, a kinematic model is established, and the real-time speed and acceleration of the logistics vehicle, i.e., the real-time path points, are input into the kinematic model to predict the trajectory envelope of the logistics vehicle within a future time window; Based on the trajectory envelope, the Monte Carlo method is used to simulate the trajectory distribution probability of the logistics vehicle under different acceleration and deceleration strategies, and the warehouse map of the smart logistics warehouse to be processed is discretized into a three-dimensional grid; According to the trajectory distribution probability of the logistics vehicle, the trajectory overlapping area is marked in the three-dimensional grid to identify the potential conflict area, and the conflict duration, relative speed and area type of the potential conflict area are calculated.

[0007] In one implementation of the present application, after identifying potential conflict areas between logistics vehicles, the method further includes: Based on the conflict severity indicators of conflict duration, relative speed, and area type, and combined with the mission residual value and current speed inertia parameters of the conflicting logistics vehicles, the priority of passage is dynamically allocated; A conflict resolution instruction set is generated based on the traffic priority and sent to the corresponding logistics vehicle; the instruction set includes a deceleration instruction, a stop and wait instruction, and a local detour instruction.

[0008] In one implementation of the present application, based on the dynamic weight vector and the conflict severity index, and using a global path replanning algorithm, an obstacle avoidance path for the logistics vehicle is generated, specifically including: Based on the global path replanning algorithm, the dynamic weight vector is integrated into the path cost function to update the cost of the road segments in the corresponding path in real time when searching for a path in the global warehouse map; For the logistics vehicles in the potential conflict area, an obstacle avoidance path with continuous acceleration is generated through polynomial interpolation technology.

[0009] In one implementation of the present application, the present invention further includes: During the local trajectory optimization process, if there is a coupling conflict between the obstacle avoidance path adjustment instructions of multiple logistics vehicles, a distributed negotiation mechanism is triggered, and the conflicting logistics vehicles’ own task attributes and path adjustment costs are broadcasted through the communication network; According to the task priority, adjustment cost and remaining path length, an auction algorithm is used to determine the winning logistics vehicle with the dominant right in conflict resolution among multiple logistics vehicles, so that the winning logistics vehicle can execute an active avoidance strategy.

[0010] In one implementation of the present application, real-time collection of execution feedback results to update reinforcement learning model parameters and dynamic weight strategies specifically includes: Monitor the mission completion time of the logistics vehicle and the energy consumption data during the mission execution, and obtain the conflict resolution results to generate a multi-dimensional performance evaluation report; the conflict resolution results include the conflict resolution success rate and the replanning trigger frequency; The actual trajectory in the execution feedback result is compared with the corresponding smooth trajectory to determine the trajectory deviation, and the trajectory deviation is used as a training sample to update the model parameters of the reinforcement learning model through a sliding time window mechanism.

[0011] In one implementation of the present application, after updating the reinforcement learning model parameters and the dynamic weight strategy, the method further includes: Conduct event annotation for conflict scenarios that require manual intervention and adjustment, extract conflict feature parameters and adjustment strategies, and build a case library of abnormal events. The abnormal event case library is loaded through a transfer learning algorithm to adjust the feature extraction parameters in the kinematic model.

[0012] On the other hand, an embodiment of the present application further provides a smart logistics scheduling device based on dynamic weights, the device comprising: at least one processor; and, a memory communicatively coupled to the at least one processor; In which, the memory stores instructions that can be executed by the at least one processor, and the instructions are executed by the at least one processor so that the at least one processor can execute the above-mentioned intelligent logistics scheduling method based on dynamic weights.

[0013] On the other hand, an embodiment of the present application also provides a non-volatile computer storage medium storing computer executable instructions. When the computer executable instructions are executed, an intelligent logistics scheduling method based on dynamic weights as described above is implemented.

[0014] The embodiments of the present application provide a method, device, and medium for intelligent logistics scheduling based on dynamic weights, which have at least the following beneficial effects: By fusing multi-source sensor data to construct a dynamic environment model and combining it with a dynamic weight adjustment mechanism driven by reinforcement learning, the problem of rigid path planning caused by the solidification of target weights in traditional logistics scheduling is effectively solved; based on spatiotemporal trajectory prediction and three-dimensional grid modeling, the spatiotemporal coupling analysis capability of conflict detection is significantly improved, avoiding the neglect of kinematic parameters and acceleration changes in traditional two-dimensional overlap detection, thereby greatly enhancing the recognition accuracy of potential conflicts and the reliability of severity assessment; through the collaborative design of global path replanning and local trajectory optimization, dynamic weight targets and dynamic constraints are taken into account in the process of obstacle avoidance path generation, achieving the optimal balance of multiple targets and ensuring smooth and executable trajectories, effectively reducing the sudden stop and start phenomenon caused by path mutations in traditional methods, thereby reducing equipment mechanical loss and energy waste; through the adaptive update of reinforcement learning models, the response capability to complex dynamic environments and emergencies is continuously improved. BRIEF DESCRIPTION OF THE DRAWINGS

[0015] The drawings described herein are used to provide a further understanding of the present application and constitute a part of the present application. The illustrative embodiments of the present application and their descriptions are used to explain the present application and do not constitute an improper limitation on the present application. In the drawings: Figure 1 A flow chart of a smart logistics scheduling method based on dynamic weights provided in an embodiment of the present application; Figure 2 A schematic diagram of the internal structure of a dynamic weight-based intelligent logistics scheduling device provided in an embodiment of the present application. DETAILED DESCRIPTION

[0016] To make the purpose, technical solutions, and advantages of this application more clear, the technical solutions of this application will be clearly and completely described below in conjunction with the specific embodiments of this application and the corresponding drawings. Obviously, the embodiments described are only part of the embodiments of this application, not all of them. Based on the embodiments in this application, all other embodiments obtained by ordinary technicians in this field without making creative efforts are within the scope of protection of this application.

[0017] The following describes in detail the technical solutions provided by various embodiments of the present application in conjunction with the accompanying drawings.

[0018] Figure 1A flow chart of a smart logistics scheduling method based on dynamic weights provided in an embodiment of the present application.

[0019] The analysis method involved in the embodiments of the present application can be implemented by a terminal device or a server, and the present application does not impose any special restrictions on this. For ease of understanding and description, the following embodiments are described in detail using a server as an example.

[0020] It should be noted that the server can be a single device or a system composed of multiple devices, that is, a distributed server, and this application does not make any specific restrictions on this.

[0021] like Figure 1 As shown, the embodiment of the present application provides a smart logistics scheduling method based on dynamic weights, including: Step 101: The operating environment data of the logistics vehicle is integrated and processed to construct a dynamic environment model. Based on the reinforcement learning model, the weight vector of each logistics parameter in the logistics vehicle is dynamically adjusted according to the real-time environmental status and multi-objective optimization requirements.

[0022] In this embodiment, the fusion processing of operating environment data is achieved through multi-source sensor processing. For example, point cloud data collected by the lidar is subjected to a voxel filtering algorithm for noise reduction, filtering out transient noise caused by dynamic obstacles such as moving people or other AGVs. It should be noted that the improved iterative closest point (ICP) algorithm updates the AGV's positioning information in real time by matching the current frame's point cloud with historical maps, ensuring the accuracy of the dynamic environment model.

[0023] As you can understand, the fusion of UWB base station signals with IMU and wheel odometry data utilizes extended Kalman filtering technology. Specifically, when UWB signal strength decreases due to occlusion, the system dynamically increases the fusion weight of the IMU data to compensate for positioning drift. For example, in densely packed warehouse areas, where UWB signals are easily obstructed, the IMU's short-term positioning stability becomes the primary reliance.

[0024] Task queue information is received in real time via the MQTT protocol. After parsing task priorities and deadlines, it is combined with the AGV's battery level and load status (e.g., whether the robotic arm is fully loaded) to construct key inputs for the dynamic environment model. Based on this, a reinforcement learning model (such as a dual deep Q network (DDQN)) generates a dynamic weight vector for the four-dimensional objective (path length, energy efficiency, task urgency, and traffic density) based on real-time traffic density, device status, and task urgency. It should be noted that weight adjustment utilizes an online policy network, and a smooth transition mechanism is implemented to avoid path oscillations caused by sudden weight changes. For example, when the battery level of an AGV falls below a preset threshold, the model automatically increases the weight for energy efficiency, guiding path planning to choose a low-power route.

[0025] Step 102: Predict the spatiotemporal trajectory of each logistics vehicle based on the kinematic model of the logistics vehicle, map the spatiotemporal trajectory into a three-dimensional grid, identify potential conflict areas between logistics vehicles, and calculate the corresponding conflict severity index.

[0026] In this embodiment, spatiotemporal trajectory prediction is based on the AGV's kinematic model. For example, the kinematic model includes parameters such as mass, maximum acceleration, and steering angle. Combined with historical velocity, acceleration sequences, and real-time path points, the trajectory envelope is predicted within a future time window. It should be noted that the trajectory envelope represents the spatial range that the AGV may occupy under different acceleration and deceleration strategies. Monte Carlo methods are used to simulate the probability distribution of multiple trajectory groups, such as simulating trajectory branches for deceleration and avoidance at intersections.

[0027] Specifically, the warehouse map is discretized into a three-dimensional space-time grid, i.e., the x-axis, y-axis, and time axis t, with each grid cell representing a local area at a specific moment. By mapping the predicted trajectory to the grid, the system marks overlapping grid cells and identifies potential conflict areas. It can be understood that the conflict severity index comprehensively calculates the conflict duration, relative speed, and area type weight. For example, in the intersection area, when two AGVs approach at high speed, the conflict severity index is significantly higher than the low-speed parallel scenario in a straight channel.

[0028] After identifying a conflict, the system dynamically prioritizes traffic based on the remaining value of the task (e.g., the value of the cargo corresponding to the higher-priority task) and the current speed and inertia parameters. Specifically, the higher-priority AGV receives the initiative, while the lower-priority AGV receives instructions to slow down or bypass. The conflict resolution instruction set includes three strategies: slow down, stop and wait, and partial bypass, ensuring flexible response to various conflict scenarios.

[0029] Step 103: Based on the dynamic weight vector and the conflict severity index, and using a global path replanning algorithm, an obstacle avoidance path for the logistics vehicle is generated, and the obstacle avoidance path is optimized through local trajectories to generate a corresponding smooth trajectory.

[0030] In this embodiment, global path replanning utilizes a modified D* Lite algorithm. For example, the path cost function incorporates a dynamic weight vector to update the cost of each road segment in real time. For example, as the task urgency weight increases, the algorithm prioritizes the shortest path; as the traffic density weight increases, congested areas are automatically avoided. It should be noted that localized high-cost road segments caused by temporary obstacles (such as dropped cargo) are dynamically marked, triggering a global path update.

[0031] For local obstacle avoidance trajectory optimization, quintic polynomial interpolation is used to generate a smooth trajectory with continuous acceleration. This algorithm avoids the sudden acceleration changes associated with traditional broken-line paths, ensuring the AGV executes its trajectory without sudden stops or starts. For example, when avoiding narrow passages, the AGV generates a curved detour trajectory, whose velocity profile is solved using a convex optimization algorithm to satisfy the maximum steering angle constraint.

[0032] When multiple AGVs' locally adjusted paths conflict, a distributed negotiation mechanism is triggered. Specifically, the conflicting AGVs broadcast their adjustment costs and their own task attributes, such as remaining task value and path inertia, over the communication network. Using an auction algorithm, they bid for priority. The winning AGV executes an active detour or deceleration instruction, while the remaining AGVs maintain their original paths. This mechanism, through distributed decision-making, reduces the computational load on the central node and improves system response efficiency.

[0033] Step 104: Execute the scheduling instructions corresponding to the smooth trajectory and collect the execution feedback results in real time to update the reinforcement learning model parameters and dynamic weight strategy.

[0034] In this embodiment, execution feedback data includes actual trajectory deviation, task completion time, energy consumption statistics, and conflict resolution success rate. It is understood that trajectory deviations are input into the reinforcement learning model via a sliding time window mechanism and used as incremental training samples to optimize the policy network parameters. For example, if the actual trajectory in a certain scenario deviates due to ground slippage, the model will adjust the weighting strategy to increase the path stability weight.

[0035] It's important to note that unusual conflict scenarios requiring manual intervention (such as sudden equipment failures) are annotated and feature parameters, such as the conflict location and AGV status, are extracted to build a case library of unusual events. Using a transfer learning algorithm to load this case library, the system fine-tunes the feature extraction layer parameters of the spatiotemporal trajectory prediction model, enhancing its ability to predict low-probability emergencies. For example, the model can quickly identify trajectory deviations caused by sudden changes in the robot arm's load and predict collision risks in advance.

[0036] Furthermore, the reinforcement learning model utilizes a prioritized experience replay mechanism, prioritizing high-conflict scenario samples for training to accelerate the model's adaptability to complex scenarios. It is understood that after real-time trajectory deviation data is combined with historical high-conflict samples, the sample distribution weights in the experience replay pool are dynamically adjusted to ensure continuous model optimization.

[0037] The above is an embodiment of the method proposed in this application. Based on the same inventive concept, this application embodiment also provides a smart logistics scheduling device based on dynamic weights, the structure of which is as follows: Figure 2 shown.

[0038] Figure 2This is a schematic diagram of the internal structure of a smart logistics scheduling device based on dynamic weights provided in an embodiment of the present application. Figure 2 As shown, the equipment includes: at least one processor; and, a memory communicatively coupled to the at least one processor; The memory stores instructions that can be executed by at least one processor, and the instructions are executed by the at least one processor to enable the at least one processor to: The operating environment data of logistics vehicles is integrated and processed to build a dynamic environment model. Based on the reinforcement learning model, the weight vectors of various logistics parameters in the logistics vehicle are dynamically adjusted according to the real-time environmental status and multi-objective optimization requirements; Predict the spatiotemporal trajectory of each logistics vehicle based on its kinematic model, map the spatiotemporal trajectory onto a three-dimensional grid, identify potential conflict areas between logistics vehicles, and calculate the corresponding conflict severity index; Based on the dynamic weight vector and conflict severity index, and using the global path replanning algorithm, the obstacle avoidance path of the logistics vehicle is generated, and the obstacle avoidance path is optimized through local trajectories to generate the corresponding smooth trajectory; Execute the scheduling instructions corresponding to the smooth trajectory and collect execution feedback results in real time to update the reinforcement learning model parameters and dynamic weight strategy.

[0039] The present application also provides a non-volatile computer storage medium storing computer-executable instructions. When the computer-executable instructions are executed, they can: The operating environment data of logistics vehicles is integrated and processed to build a dynamic environment model. Based on the reinforcement learning model, the weight vectors of various logistics parameters in the logistics vehicle are dynamically adjusted according to the real-time environmental status and multi-objective optimization requirements; Predict the spatiotemporal trajectory of each logistics vehicle based on its kinematic model, map the spatiotemporal trajectory onto a three-dimensional grid, identify potential conflict areas between logistics vehicles, and calculate the corresponding conflict severity index; Based on the dynamic weight vector and conflict severity index, and using the global path replanning algorithm, the obstacle avoidance path of the logistics vehicle is generated, and the obstacle avoidance path is optimized through local trajectories to generate the corresponding smooth trajectory; Execute the scheduling instructions corresponding to the smooth trajectory and collect execution feedback results in real time to update the reinforcement learning model parameters and dynamic weight strategy.

[0040] The various embodiments in this application are described in a progressive manner. Similar portions between the various embodiments can be referred to in conjunction with each other. Each embodiment focuses on the differences between the other embodiments. In particular, the device and medium embodiments are generally similar to the method embodiments, so their descriptions are relatively simple. For relevant portions, refer to the descriptions of the method embodiments.

[0041] The devices and media provided in the embodiments of the present application correspond one-to-one to the methods. Therefore, the devices and media also have similar beneficial technical effects to their corresponding methods. Since the beneficial technical effects of the methods have been described in detail above, the beneficial technical effects of the devices and media will not be repeated here.

[0042] Those skilled in the art will appreciate that the embodiments of the present application may be provided as methods, systems, or computer program products. Therefore, the present application may take the form of an entirely hardware embodiment, an entirely software embodiment, or an embodiment combining software and hardware. Furthermore, the present application may take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to magnetic disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0043] The present application is described with reference to the flowcharts and / or block diagrams of the methods, devices (systems), and computer program products according to the embodiments of the present application. It should be understood that each process and / or block in the flowchart and / or block diagram, as well as the combination of processes and / or blocks in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing device to produce a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the processes in the flowchart and / or block diagram. Figure 1 a process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.

[0044] These computer program instructions may also be stored in a computer readable memory that can direct a computer or other programmable data processing device to work in a specific manner, so that the instructions stored in the computer readable memory produce an article of manufacture comprising an instruction device, which implements the process Figure 1 a process or multiple processes and / or boxes Figure 1 The function specified in one or more boxes.

[0045] These computer program instructions can also be loaded onto a computer or other programmable data processing device so that a series of operational steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing the instructions executed on the computer or other programmable device for implementing the process. Figure 1 a process or multiple processes and / or boxes Figure 1 A step that specifies a function in one or more boxes.

[0046] In a typical configuration, a computing device includes one or more processors (CPUs), input / output interfaces, network interfaces, and memory.

[0047] Memory may include non-permanent storage in a computer-readable medium, in the form of random access memory (RAM) and / or non-volatile memory, such as read-only memory (ROM) or flash RAM. Memory is an example of a computer-readable medium.

[0048] Computer-readable media includes both permanent and non-permanent, removable and non-removable media that can be implemented using any method or technology for information storage. Information can be computer-readable instructions, data structures, program modules, or other data. Examples of computer storage media include, but are not limited to, phase-change RAM (PRAM), static random access memory (SRAM), dynamic random access memory (DRAM), other types of random access memory (RAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), flash memory or other memory technologies, compact disc read-only memory (CD-ROM), digital versatile disc (DVD) or other optical storage, magnetic cassettes, magnetic disk storage or other magnetic storage devices, or any other non-transmission media that can be used to store information that can be accessed by a computing device. As defined herein, computer-readable media does not include transitory computer-readable media, such as modulated data signals and carrier waves.

[0049] It should also be noted that the terms "comprises," "includes," or any other variations thereof are intended to encompass non-exclusive inclusion, such that a process, method, commodity, or apparatus that includes a series of elements includes not only those elements but also other elements not explicitly listed, or includes elements inherent to such process, method, commodity, or apparatus. In the absence of further limitations, an element defined by the phrase "comprises a ..." does not exclude the presence of other identical elements in the process, method, commodity, or apparatus that includes the element.

[0050] The above are merely embodiments of the present application and are not intended to limit the present application. For those skilled in the art, the present application may have various changes and variations. Any modifications, equivalent replacements, improvements, etc. made within the spirit and principles of the present application should all be included within the scope of the claims of the present application.

Claims

1. A smart logistics scheduling method based on dynamic weights, characterized in that: The method comprises: The operating environment data of the logistics vehicle is integrated and processed to construct a dynamic environment model. Based on the reinforcement learning model, the weight vector of each logistics parameter in the logistics vehicle is dynamically adjusted according to the real-time environmental status and multi-objective optimization requirements; Predicting the spatiotemporal trajectory of each logistics vehicle based on its kinematic model, mapping the spatiotemporal trajectory to a three-dimensional grid, identifying potential conflict areas between logistics vehicles, and calculating corresponding conflict severity indicators; Based on the dynamic weight vector and the conflict severity index, and using a global path replanning algorithm, an obstacle avoidance path for the logistics vehicle is generated, and the obstacle avoidance path is optimized through local trajectories to generate a corresponding smooth trajectory; Execute the scheduling instructions corresponding to the smooth trajectory and collect execution feedback results in real time to update the reinforcement learning model parameters and dynamic weight strategy.

2. The intelligent logistics scheduling method based on dynamic weight according to claim 1 is characterized in that: The operation environment data of logistics vehicles is integrated and processed, including: Collecting real-time operating environment data of logistics vehicles through multi-source sensors; the operating environment data includes lidar point cloud data, ultra-wideband base station signals, equipment status parameters and task queue information; De-noising the LiDAR point cloud data based on a voxel filtering algorithm, and matching the historical point cloud map with an iterative closest point algorithm to update the positioning information of the logistics vehicle in real time; fusing the ultra-wideband base station signal with inertial measurement unit data and wheel odometer information in the device state parameters, constructing a three-dimensional network through extended Kalman filtering, and dynamically adjusting the fusion weight according to signal strength; The task queue information is received in real time through a message queue telemetry transmission protocol to parse the task priority and deadline in the task queue information and monitor the battery power and load status of the logistics vehicle.

3. The intelligent logistics scheduling method based on dynamic weight according to claim 1 is characterized in that: Based on the kinematic model of logistics vehicles, the spatiotemporal trajectory of each logistics vehicle is predicted, and the spatiotemporal trajectory is mapped to a three-dimensional grid. Potential conflict areas between logistics vehicles are identified, and the corresponding conflict severity index is calculated, including: Based on the historical speed, historical acceleration, and path point sequence of the logistics vehicle, a kinematic model is established, and the real-time speed and acceleration of the logistics vehicle, i.e., the real-time path points, are input into the kinematic model to predict the trajectory envelope of the logistics vehicle within a future time window; Based on the trajectory envelope, the Monte Carlo method is used to simulate the trajectory distribution probability of the logistics vehicle under different acceleration and deceleration strategies, and the warehouse map of the smart logistics warehouse to be processed is discretized into a three-dimensional grid; According to the trajectory distribution probability of the logistics vehicle, the trajectory overlapping area is marked in the three-dimensional grid to identify the potential conflict area, and the conflict duration, relative speed and area type of the potential conflict area are calculated.

4. The intelligent logistics scheduling method based on dynamic weight according to claim 3 is characterized in that: After identifying potential conflict areas between logistics vehicles, the method further includes: Based on the conflict severity indicators of conflict duration, relative speed, and area type, and combined with the mission residual value and current speed inertia parameters of the conflicting logistics vehicles, the priority of passage is dynamically allocated; A conflict resolution instruction set is generated based on the traffic priority and sent to the corresponding logistics vehicle; the instruction set includes a deceleration instruction, a stop and wait instruction, and a local detour instruction.

5. The intelligent logistics scheduling method based on dynamic weight according to claim 1 is characterized in that: Based on the dynamic weight vector and the conflict severity index, and using a global path replanning algorithm, an obstacle avoidance path for the logistics vehicle is generated, specifically including: Based on the global path replanning algorithm, the dynamic weight vector is integrated into the path cost function to update the cost of the road segments in the corresponding path in real time when searching for a path in the global warehouse map; For the logistics vehicles in the potential conflict area, an obstacle avoidance path with continuous acceleration is generated through polynomial interpolation technology.

6. The intelligent logistics scheduling method based on dynamic weight according to claim 5 is characterized in that: The method further comprises: During the local trajectory optimization process, if there is a coupling conflict between the obstacle avoidance path adjustment instructions of multiple logistics vehicles, a distributed negotiation mechanism is triggered, and the conflicting logistics vehicles’ own task attributes and path adjustment costs are broadcasted through the communication network; According to the task priority, adjustment cost and remaining path length, an auction algorithm is used to determine the winning logistics vehicle with the dominant right in conflict resolution among multiple logistics vehicles, so that the winning logistics vehicle can execute an active avoidance strategy.

7. The intelligent logistics scheduling method based on dynamic weight according to claim 1 is characterized in that: Collect execution feedback results in real time to update reinforcement learning model parameters and dynamic weight strategies, including: Monitor the mission completion time of the logistics vehicle and the energy consumption data during the mission execution, and obtain the conflict resolution results to generate a multi-dimensional performance evaluation report; the conflict resolution results include the conflict resolution success rate and the replanning trigger frequency; The actual trajectory in the execution feedback result is compared with the corresponding smooth trajectory to determine the trajectory deviation, and the trajectory deviation is used as a training sample to update the model parameters of the reinforcement learning model through a sliding time window mechanism.

8. The intelligent logistics scheduling method based on dynamic weight according to claim 7 is characterized in that: After updating the reinforcement learning model parameters and the dynamic weight strategy, the method further includes: Conduct event annotation for conflict scenarios that require manual intervention and adjustment, extract conflict feature parameters and adjustment strategies, and build a case library of abnormal events. The abnormal event case library is loaded through a transfer learning algorithm to adjust the feature extraction parameters in the kinematic model.

9. A smart logistics scheduling device based on dynamic weights, characterized in that: The device comprises: at least one processor; and, a memory communicatively coupled to the at least one processor; In which, the memory stores instructions that can be executed by the at least one processor, and the instructions are executed by the at least one processor so that the at least one processor can execute a smart logistics scheduling method based on dynamic weights as described in any one of claims 1-8.

10. A non-volatile computer storage medium storing computer executable instructions, characterized in that: When the computer executable instructions are executed, a smart logistics scheduling method based on dynamic weights as described in any one of claims 1 to 8 is implemented.

Images