Method for collaborative scheduling of AGV and automated warehouse
By employing data acquisition and prediction models, visual recognition and dynamic posture adjustment, and global path planning, the problem of coordinated scheduling between AGVs and automated warehouses has been solved, achieving efficient and stable cargo flow.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- GUANGZHOU HOTENT SOFTWARE CO LTD
- Filing Date
- 2026-04-20
- Publication Date
- 2026-07-14
AI Technical Summary
Existing technologies for the collaborative scheduling of AGVs and automated warehouses suffer from problems such as time mismatch, reliance on fixed positioning devices for position correction, and lack of global coordination in multi-AGV path planning. These issues lead to waiting, missed docking opportunities, positional deviations, channel conflicts, and low passage efficiency.
By employing methods such as data acquisition and data pool construction, task prediction and window delineation, deviation identification and attitude adjustment, multi-machine collaborative path planning, driving control and conflict avoidance, efficiency statistics and strategy optimization, precise docking and efficient collaboration between AGVs and automated three-dimensional warehouses can be achieved.
It improves the collaborative scheduling efficiency of AGVs and automated warehouses, ensures the accuracy and safety of goods retrieval and placement, avoids traffic conflicts, optimizes the traffic order of multi-vehicle collaboration, and continuously iterates and optimizes scheduling strategies to maintain a highly efficient and stable operating state.
Smart Images
Figure CN122387045A_ABST
Abstract
Description
Technical Field
[0001] This invention relates to the field of intelligent warehousing technology, specifically a method for collaborative scheduling of AGVs and automated storage and retrieval systems. Background Technology
[0002] In modern intelligent warehousing systems, the coordinated scheduling of AGVs and automated storage and retrieval systems (AS / RS) is a core element in improving cargo flow efficiency and reducing operating costs, directly impacting a company's supply chain responsiveness and market competitiveness. With the rapid development of e-commerce, intelligent manufacturing, and other industries, warehousing operations are experiencing explosive growth, placing increasingly stringent demands on the efficiency and accuracy of their coordination.
[0003] In existing technologies, the handling efficiency of AGVs is highly dependent on the seamless connection of automated warehouses: on the one hand, AGVs need to arrive at the warehouse entrance and exit at a specified time to complete the picking and placing of goods. If the time when the outbound task arrives at the exit does not match the arrival time of the AGV, it will cause the AGV to wait for a long time or miss the docking opportunity; on the other hand, AGVs need to be accurately adapted to the location of the entrance and exit, otherwise it will affect the safety and efficiency of picking and placing goods. The lack of an accurate prediction mechanism for outbound task completion time leads to a disconnect between AGVs and automated warehouse operations. Furthermore, the positional correction of AGVs during goods handling relies heavily on fixed positioning devices, lacking dynamic recognition and adjustment capabilities, making it difficult to adapt to positional deviations in complex scenarios. Additionally, multi-AGV path planning prioritizes the shortest path for a single vehicle without considering global task coordination, easily causing channel conflicts and congestion, reducing overall traffic efficiency, and failing to meet the operational requirements of intelligent warehousing. Therefore, a method for collaborative scheduling of AGVs and automated warehouses is proposed. Summary of the Invention
[0004] To address the shortcomings of existing technologies, this invention provides a method for collaborative scheduling of AGVs and automated warehouses, thereby solving the technical problems of insufficient coordination, connection, passage, and optimization in collaborative scheduling.
[0005] To achieve the above objectives, the present invention provides the following technical solution: a method for collaborative scheduling of AGVs and automated storage and retrieval systems, comprising the following steps: S1 Data Acquisition and Data Pool Construction: First, the coordinates of the inbound and outbound entrances and the aisle layout parameters of the automated warehouse are collected. Then, the real-time position, speed and power information are obtained through the AGV positioning unit, and the inbound and outbound task data of the warehouse management system are received simultaneously. A basic data pool is established based on the data obtained above. S2 Task Prediction and Window Delineation: Secondly, based on the historical mission duration, the real-time status of the stacker crane, and real-time traffic congestion feedback information, a time-series prediction model is used to predict the time when goods arrive at the target inlet / outlet, and the docking time window is determined by combining the AGV distance and road conditions. S3 Deviation Identification and Attitude Adjustment: After the AGV travels to the target port buffer area, the vision recognition module is activated to detect the position deviation, and the vehicle body and the loading and unloading device are dynamically adjusted through the drive system to complete the docking preparation; S4 Multi-Machine Collaborative Path Planning: Subsequently, the scheduling system summarizes the status of multiple AGVs, docking windows, and multi-dimensional task priority weighted results, and uses a global path planning algorithm that incorporates conflict avoidance and a dynamic right-of-way allocation algorithm to plan cooperative driving paths and mark conflict nodes and avoidance areas. S5 Driving Control and Collision Avoidance: The AGV travels along the planned path in step S4, matching the travel progress with the docking time window in real time, and adjusting the speed according to safety rules to match the docking opportunity, while following the cooperative path to avoid traffic conflicts. S6 Efficiency Statistics and Strategy Optimization: The system provides real-time statistics on the success rate of single-vehicle docking and the efficiency of multi-vehicle collaborative passage, and compares these results with preset targets. If the targets are not met, the system returns to adjust the prediction model parameters or path planning strategy.
[0006] Preferably, in step S1, when collecting the coordinates of the warehouse entrance and exit and the channel layout parameters, various basic information is simultaneously classified, sorted, and standardized. After the AGV positioning unit obtains real-time status information, data verification and anomaly filtering are simultaneously completed. When receiving task data from the warehouse management system, the task batch and operation scenario are simultaneously associated. This allows for the classification, sorting, and standardization of basic information, data verification and anomaly filtering, and association of task batches and operation scenarios, making the basic data more organized and reliable, and providing accurate and practical data support for subsequent scheduling.
[0007] Preferably, the time-series prediction model used in step S2 is equipped with an adaptive learning program for warehousing operation scenarios. This allows the prediction logic to be autonomously optimized based on historical task characteristics of different operation periods and different types of goods. It also dynamically corrects predictions by combining real-time traffic congestion feedback information and the real-time operation status of the stacker crane. Furthermore, it synchronously matches on-site operational emergencies and AGV driving conditions when defining the docking time window. This enables the prediction model to have scenario adaptive learning capabilities, dynamically correct prediction results by combining real-time information, and match on-site emergencies and driving conditions within the time window, significantly improving prediction accuracy and making the connection between AGV and automated warehouse operations smoother.
[0008] Preferably, in step S3, after the AGV enters the target inlet / outlet buffer area, while activating the vision recognition module, multimodal joint deviation detection is performed. This involves comprehensively identifying the relative positional deviations between the vehicle body, the pick-and-place device, the inlet / outlet, and the goods. Subsequently, the drive system performs coordinated dynamic adjustment based on the deviation detection results, simultaneously adjusting the vehicle body's driving posture and the pick-and-place device's operating posture. It also autonomously compensates for environmental interference. By comprehensively identifying positional deviations through multimodal joint detection, and by the drive system collaboratively adjusting posture and autonomously compensating for environmental interference, the accuracy of deviation identification and posture adjustment is effectively improved, ensuring the stability and reliability of docking preparation.
[0009] Preferably, in step S3, after completing the position deviation detection and attitude adjustment, the docking preparation verification process is automatically triggered. That is, the visual recognition module confirms for the second time whether the attitude of the vehicle body and the pick-up and place device meets the docking standard. If it does not meet the standard, the cyclic fine-tuning mechanism is started until the docking requirements are met. By confirming the docking attitude through the second verification, the cyclic fine-tuning is started when the standard is not met, which can completely eliminate the docking deviation and ensure that the vehicle body and the pick-up and place device fully meet the docking standard, significantly improving the success rate and stability of single vehicle docking.
[0010] Preferably, in step S4, when the scheduling system performs weighted calculations on the multi-dimensional task priorities, it simultaneously combines task attributes, work scenarios, AGV status, and warehouse operation load for comprehensive analysis and dynamically updates the task priority ranking. At the same time, it uses a global path planning algorithm to predict potential traffic conflicts in advance and coordinates with a dynamic right-of-way allocation algorithm to allocate permissions in different time zones based on conflict levels and task priorities. By comprehensively analyzing multiple factors and dynamically updating task priorities, combined with conflict prediction and time-sharing right-of-way allocation, the scheduling of multiple AGVs becomes more reasonable, effectively ensuring the execution of high-priority tasks and improving the orderliness of collaborative operations.
[0011] Preferably, in step S4, when planning the collaborative driving path of multiple AGVs, a hierarchical global planning mode is adopted. First, the main traffic path is determined, and then the branch traffic routes are refined. In addition, a multi-level avoidance scheme is formulated for the marked conflict nodes, and the avoidance area is dynamically adjusted according to the real-time distribution of AGVs and the progress of the operation. At the same time, the path and permission information are synchronously sent to each AGV. By adopting hierarchical path planning, combined with multi-level avoidance schemes and dynamic avoidance areas, traffic conflicts can be comprehensively avoided, and the path and permission information can be synchronously sent to achieve seamless connection between scheduling instructions and AGV execution.
[0012] Preferably, in step S5, when the AGV adjusts its speed to match the docking time window, it simultaneously analyzes the work efficiency and driving energy consumption, autonomously selects the optimal driving mode within the framework of safety rules, and actively responds to the real-time control instructions of the scheduling system while following the collaborative path to avoid conflicts. At the same time, it synchronously feeds back its own driving progress and attitude status to the scheduling system. By allowing the AGV to take into account both work efficiency and driving energy consumption, autonomously select the optimal driving mode, actively respond to scheduling and feed back driving status, the intelligence and energy efficiency of scheduling are improved while ensuring docking efficiency.
[0013] Preferably, in step S5, during the AGV's journey along the planned path, the matching degree between its own travel progress and the docking time window is continuously calibrated. If path deviation or travel delay occurs, the AGV autonomously initiates path correction and speed adaptation adjustment. Under the premise of avoiding traffic conflicts, priority is given to ensuring that the work preparation is completed within the docking time window. This allows the AGV to continuously calibrate the matching degree between its travel progress and the time window, and autonomously correct the path and speed in case of abnormalities. On the basis of avoiding conflicts, it ensures that the preparation is completed within the docking window, thus ensuring the timeliness of the operation.
[0014] Preferably, in step S6, when calculating the success rate of single-vehicle docking and the efficiency of multi-vehicle collaborative passage, the execution details of each operation are first fully recorded. Then, the statistical results are compared with the preset optimization targets item by item. When the preset targets are not met, the parameters of the time-series prediction model or the optimization path planning strategy are adjusted in a targeted manner to form a continuously iterative collaborative scheduling optimization closed loop. The operation details are fully recorded and the optimization targets are compared item by item. After accurately locating the problem, the strategy is adjusted in a targeted manner to form a closed-loop iterative optimization mechanism, continuously improving the scheduling effect and ensuring that collaborative operations remain efficient and stable in the long term.
[0015] Compared with existing technologies, this invention provides a method for collaborative scheduling of AGVs and automated warehouses, which has the following beneficial effects: This method for collaborative scheduling of AGVs and automated storage and retrieval systems (AS / RS) integrates AS / RS layout parameters, AGV real-time status, and warehouse task data to construct a basic data pool. This provides a unified and complete data basis for end-to-end scheduling, ensuring precise support for scheduling instructions at each stage. Furthermore, by relying on historical task durations, stacker crane real-time status, and real-time road conditions, task prediction and docking time windows are defined, allowing AGVs to precisely match the AS / RS operation rhythm and avoiding waiting or delays caused by operational gaps. Moreover, by leveraging visual recognition bias and dynamic posture adjustment, docking position errors can be eliminated, ensuring the accuracy and safety of goods retrieval and placement, and improving the stability of single-vehicle docking. The use of global path planning combined with dynamic right-of-way allocation can coordinate the travel routes of multiple AGVs, proactively avoiding traffic conflicts and congestion, optimizing the traffic order of multi-vehicle collaboration. Simultaneously, by real-time statistics of operational efficiency and iterative optimization of scheduling strategies, the predictive model and path scheme can be continuously improved, continuously enhancing the collaborative scheduling effect and maintaining a stable and efficient operating state in the long term, thereby improving the overall efficiency and quality of warehouse goods flow. Attached Figure Description
[0016] Figure 1 This is a flowchart of the method of the present invention. Detailed Implementation
[0017] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.
[0018] Please see Figure 1 This invention provides a technical solution, a method for collaborative scheduling of AGVs and automated warehouses, comprising the following steps: S1 Data Acquisition and Data Pool Construction: First, the coordinates of the inbound and outbound entrances and the aisle layout parameters of the automated warehouse are collected. Then, the real-time position, speed and power information are obtained through the AGV positioning unit, and the inbound and outbound task data of the warehouse management system are received simultaneously. A basic data pool is established based on the data obtained above. When collecting the coordinates of the entrance and exit of the automated warehouse and the layout parameters of the passage, various basic information is classified, sorted and standardized simultaneously. After the AGV positioning unit obtains the real-time status information, data verification and anomaly filtering are completed simultaneously. When receiving task data from the warehouse management system, the batch to which the task belongs and the operation scenario are associated simultaneously. S2 Task Prediction and Window Delineation: Secondly, based on the historical mission duration, the real-time status of the stacker crane, and real-time traffic congestion feedback information, a time-series prediction model is used to predict the time when goods arrive at the target inlet / outlet, and the docking time window is determined by combining the AGV distance and road conditions. Furthermore, the time-series prediction model is equipped with an adaptive learning program for warehouse operation scenarios, which enables the prediction logic to be automatically optimized based on the historical task characteristics of different operation periods and different types of goods. It also dynamically corrects the prediction logic by combining real-time traffic congestion feedback information and the real-time operation status of the stacker crane. Moreover, it simultaneously matches on-site operation emergencies and AGV driving conditions when defining docking time windows. S3 Deviation Identification and Attitude Adjustment: After the AGV travels to the target port buffer area, the vision recognition module is activated to detect the position deviation, and the vehicle body and the loading and unloading device are dynamically adjusted through the drive system to complete the docking preparation; After the AGV enters the buffer area of the target inlet / outlet, it performs multimodal joint deviation detection while activating the vision recognition module. This means that it identifies the relative positional deviations between the vehicle body, the pick-and-place device, the inlet / outlet, and the goods in all directions. Then, the drive system performs collaborative dynamic adjustment based on the deviation detection results, simultaneously adjusting the vehicle body's driving posture and the pick-and-place device's operating posture, while also autonomously compensating for environmental interference on site. After completing the position deviation detection and attitude adjustment, the docking preparation verification process is automatically triggered. That is, the visual recognition module confirms for the second time whether the attitude of the vehicle body and the pick-up and place device meets the docking standard. If it does not meet the standard, the cyclic fine-tuning mechanism is started until the docking requirements are met. S4 Multi-Machine Collaborative Path Planning: Subsequently, the scheduling system summarizes the status of multiple AGVs, docking windows, and multi-dimensional task priority weighted results, and uses a global path planning algorithm that incorporates conflict avoidance and a dynamic right-of-way allocation algorithm to plan cooperative driving paths and mark conflict nodes and avoidance areas. When the scheduling system performs weighted calculations on the priority of multi-dimensional tasks, it simultaneously combines task attributes, work scenarios, AGV status and warehouse operation load for comprehensive analysis and dynamically updates task priority ranking. At the same time, it uses a global path planning algorithm to predict potential traffic conflicts in advance and uses a dynamic right-of-way allocation algorithm to allocate permissions in different time zones based on conflict level and task priority. When planning the collaborative driving path of multiple AGVs, a hierarchical global planning mode is adopted. First, the main travel path is determined, then the branch travel routes are refined, and multi-level avoidance schemes are formulated for marked conflict nodes. The avoidance area is dynamically adjusted according to the real-time distribution of AGVs on site and the progress of the operation. At the same time, the path and permission information are synchronously distributed to each AGV. S5 Driving Control and Collision Avoidance: The AGV travels along the planned path in step S4, matching the travel progress with the docking time window in real time, and adjusting the speed according to safety rules to match the docking opportunity, while following the cooperative path to avoid traffic conflicts. When the AGV adjusts its speed to match the docking time window, it simultaneously analyzes the work efficiency and driving energy consumption, and autonomously selects the optimal driving mode within the framework of safety rules. When following the collaborative path to avoid conflicts, it actively responds to the real-time control instructions of the scheduling system, and simultaneously feeds back its own driving progress and attitude status to the scheduling system. As the AGV travels along the planned path, it continuously calibrates the matching degree between its own travel progress and the docking time window. If there is a path deviation or travel delay, it will autonomously initiate path correction and speed adaptation adjustment. Under the premise of avoiding traffic conflicts, it will prioritize ensuring that the work preparation is completed within the docking time window. S6 Efficiency Statistics and Strategy Optimization: Real-time statistics are collected on the success rate of single-vehicle docking and the efficiency of multi-vehicle collaborative passage, and compared with the preset target. If the target is not met, the prediction model parameters or path planning strategy are adjusted. When calculating the success rate of single-vehicle docking and the efficiency of multi-vehicle collaborative passage, the execution details of each operation are first fully recorded. Then, the statistical results are compared with the preset optimization targets item by item. When the preset targets are not met, the parameters of the time series prediction model or the path planning strategy are adjusted accordingly to form a continuously iterative collaborative scheduling optimization closed loop.
[0019] This solution integrates automated warehouse layout parameters, AGV real-time status, and warehouse task data to construct a basic data pool, providing a unified and complete data basis for end-to-end scheduling. This ensures precise support for scheduling instructions at each stage. Furthermore, by leveraging historical task durations, stacker crane real-time status, and real-time road conditions, it predicts tasks and defines docking time windows, allowing AGVs to precisely match the rhythm of automated warehouse operations and avoiding waiting or delays caused by operational gaps. Moreover, by utilizing visual recognition bias and dynamic posture adjustment, it eliminates docking position errors, ensuring the accuracy and safety of goods retrieval and placement, and improving the stability of single-vehicle docking. The use of global path planning combined with dynamic right-of-way allocation coordinates the routes of multiple AGVs, proactively avoiding traffic conflicts and congestion, optimizing the traffic order of multi-vehicle collaboration. Simultaneously, by real-time statistics of operational efficiency and iterative optimization of scheduling strategies, it continuously refines predictive models and path schemes, constantly improving collaborative scheduling effects and maintaining a stable and efficient operational state in the long term, thereby improving the overall efficiency and quality of warehouse goods flow.
[0020] It should be noted that, in this document, relational terms such as "first" and "second" are used only to distinguish one entity or operation from another, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Furthermore, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such process, method, article, or apparatus.
[0021] Although embodiments of the invention have been shown and described, it will be understood by those skilled in the art that various changes, modifications, substitutions and alterations can be made to these embodiments without departing from the principles and spirit of the invention, the scope of which is defined by the appended claims and their equivalents.
Claims
1. A method for collaborative scheduling of AGVs and automated storage and retrieval systems, characterized in that, Includes the following steps: S1 Data Acquisition and Data Pool Construction: First, the coordinates of the inbound and outbound entrances and the aisle layout parameters of the automated warehouse are collected. Then, the real-time position, speed and power information are obtained through the AGV positioning unit, and the inbound and outbound task data of the warehouse management system are received simultaneously. A basic data pool is established based on the data obtained above. S2 Task Prediction and Window Delineation: Secondly, based on the historical mission duration, the real-time status of the stacker crane, and real-time traffic congestion feedback information, a time-series prediction model is used to predict the time when goods arrive at the target inlet / outlet, and the docking time window is determined by combining the AGV distance and road conditions. S3 Deviation Identification and Attitude Adjustment: After the AGV travels to the target port buffer area, the vision recognition module is activated to detect the position deviation, and the vehicle body and the loading and unloading device are dynamically adjusted through the drive system to complete the docking preparation; S4 Multi-Machine Collaborative Path Planning: Subsequently, the scheduling system summarizes the status of multiple AGVs, docking windows, and multi-dimensional task priority weighted results, and uses a global path planning algorithm that incorporates conflict avoidance and a dynamic right-of-way allocation algorithm to plan cooperative driving paths and mark conflict nodes and avoidance areas. S5 Driving Control and Collision Avoidance: The AGV travels along the planned path in step S4, matching the travel progress with the docking time window in real time, and adjusting the speed according to safety rules to match the docking opportunity, while following the cooperative path to avoid traffic conflicts. S6 Efficiency Statistics and Strategy Optimization: The system provides real-time statistics on the success rate of single-vehicle docking and the efficiency of multi-vehicle collaborative passage, and compares these results with preset targets. If the targets are not met, the system returns to adjust the prediction model parameters or path planning strategy.
2. The method for collaborative scheduling of AGVs and automated warehouses according to claim 1, characterized in that: In step S1, when collecting the coordinates of the entrance and exit of the automated warehouse and the channel layout parameters, various basic information is classified, sorted and standardized simultaneously. After the AGV positioning unit obtains real-time status information, data verification and anomaly filtering are completed simultaneously. When receiving task data from the warehouse management system, the batch to which the task belongs and the operation scenario are associated simultaneously.
3. The method for collaborative scheduling of AGVs and automated warehouses according to claim 1, characterized in that: In step S2, the time-series prediction model is equipped with an adaptive learning program for warehousing operation scenarios, which enables the prediction logic to be automatically optimized based on the historical task characteristics of different operation periods and different types of goods. It also dynamically corrects the prediction logic by combining real-time traffic congestion feedback information and the real-time operation status of the stacker crane, and simultaneously matches the on-site operation emergencies and AGV driving conditions when defining the docking time window.
4. The method for collaborative scheduling of AGVs and automated storage and retrieval systems according to claim 1, characterized in that: In step S3, after the AGV enters the buffer area of the target inlet / outlet, it performs multimodal joint deviation detection while activating the vision recognition module. This means that it identifies the relative positional deviations between the vehicle body, the pick-and-place device, the inlet / outlet, and the goods in all directions. Subsequently, the drive system performs collaborative dynamic adjustment based on the deviation detection results, simultaneously adjusting the vehicle body's driving posture and the pick-and-place device's operating posture, while also autonomously compensating for environmental interference on site.
5. The method for collaborative scheduling of AGVs and automated warehouses according to claim 1, characterized in that: In step S3, after completing the position deviation detection and attitude adjustment, the docking preparation verification process is automatically triggered. That is, the visual recognition module confirms for the second time whether the attitude of the vehicle body and the pick-up and place device meets the docking standard. If the standard is not met, the cyclic fine-tuning mechanism is started until the docking requirements are met.
6. The method for collaborative scheduling of AGVs and automated storage and retrieval systems according to claim 1, characterized in that: In step S4, when the scheduling system performs weighted calculation of multi-dimensional task priorities, it simultaneously combines task attributes, work scenarios, AGV status and warehouse operation load for comprehensive analysis and dynamically updates task priority ranking. At the same time, it uses a global path planning algorithm to predict potential passage conflicts in advance and uses a dynamic passage right allocation algorithm to allocate permissions in different time zones based on conflict level and task priority.
7. The method for collaborative scheduling of AGVs and automated storage and retrieval systems according to claim 1, characterized in that: In step S4, when planning the collaborative driving path of multiple AGVs, a hierarchical global planning mode is adopted. First, the main driving path is determined, then the branch driving routes are refined, and a multi-level avoidance scheme is formulated for the marked conflict nodes. The avoidance area is dynamically adjusted according to the real-time distribution of AGVs on site and the progress of the operation. At the same time, the path and permission information are synchronously distributed to each AGV.
8. The method for collaborative scheduling of AGVs and automated warehouses according to claim 1, characterized in that: In step S5, when the AGV adjusts its speed to match the docking time window, it simultaneously analyzes the operation timeliness and driving energy consumption, autonomously selects the optimal driving mode within the framework of safety rules, and actively responds to the real-time control instructions of the scheduling system while following the collaborative path to avoid conflicts. At the same time, it synchronously feeds back its own driving progress and attitude status to the scheduling system.
9. The method for collaborative scheduling of AGVs and automated storage and retrieval systems according to claim 1, characterized in that: In step S5, as the AGV travels along the planned path, it continuously calibrates the matching degree between its own travel progress and the docking time window. If there is a path deviation or travel delay, it will automatically start path correction and speed adaptation adjustment. Under the premise of avoiding traffic conflicts, it will prioritize ensuring that the work preparation is completed within the docking time window.
10. The method for collaborative scheduling of AGVs and automated storage and retrieval systems according to claim 1, characterized in that: In step S6, when calculating the success rate of single-vehicle docking and the efficiency of multi-vehicle collaborative passage, the execution details of each operation are first fully recorded. Then, the statistical results are compared with the preset optimization targets item by item. When the preset targets are not met, the parameters of the time-series prediction model or the path planning strategy are adjusted in a targeted manner to form a continuously iterative collaborative scheduling optimization closed loop.