Warehouse scheduling method based on WMS and AGV cooperation

Abstract

The invention relates to the technical field of warehouse logistics, and discloses a warehouse scheduling method based on WMS and AGV cooperation. The method comprises the following steps: constructing a dynamic storage location data center, and integrating storage location occupation states, material turnover rates and other parameters; generating a dynamic storage location allocation scheme by using a multi-factor decision algorithm; according to the real-time warehouse-in and warehouse-out instruction, an AGV carrying task chain is generated according to a task decomposition rule; constructing a multi-AGV cooperative path planning network by using the space-time map; inputting the storage location allocation scheme and the path planning network into the coupling scheduling model to generate an AGV cluster scheduling instruction set; and an AGV driving track is adjusted through an anti-collision optimization algorithm, and a final scheduling strategy is output. According to the invention, intelligentization and high efficiency of storage scheduling are realized, the storage space utilization rate, the operation efficiency and the safety are improved, the operation cost is reduced, and the method is suitable for various storage logistics scenes.

Description

Technical Field

[0001] The present invention relates to the field of warehousing and logistics technology, and in particular to a warehousing scheduling method based on collaboration between WMS and AGV. Background Art

[0002] In the modern warehousing and logistics industry, with the rapid development of e-commerce, manufacturing, and other sectors, storage scale continues to expand, and the variety and frequency of goods entering and leaving the warehouse have increased significantly. Traditional warehouse scheduling methods face numerous challenges. Early warehouse management relied heavily on manual experience for location allocation and task scheduling, which was inefficient and prone to errors. Manual recording of location information made it difficult to track location occupancy in real time, making it difficult to locate materials, resulting in long inbound and outbound times and low storage space utilization. For example, in large e-commerce warehouses, cargo volumes surge during promotional events, and manual scheduling was unable to respond in a timely manner, leading to severe backlogs.

[0003] With technological advancements, some warehouses have introduced warehouse management systems (WMS), enabling information-based management of storage locations and inventory, which has significantly improved warehouse management efficiency. However, WMS often allocate storage locations based on simple rules, such as first-in, first-out (FIFO) and fixed partitions, without fully considering key factors such as material turnover, storage environment requirements, and inbound and outbound frequency. This can result in materials being placed in inappropriate locations, increasing transportation distances and time costs.

[0004] Automated Guided Vehicles (AGVs) are increasingly being used in warehousing, enabling automated handling and reducing labor costs. However, when multiple AGVs are working together, path planning becomes a significant issue. Existing path planning methods cannot accurately predict path conflicts, leading to frequent collisions and congestion between AGVs, reducing overall handling efficiency. For example, in high-density warehouses with a large number of AGVs, traditional path planning struggles to effectively coordinate them. This often results in multiple vehicles competing for space, resulting in extended downtime.

[0005] Furthermore, there was a lack of effective coordination between the WMS and AGVs. Data exchange between the two was poor, and information updates were untimely, preventing warehouse scheduling from forming an integrated whole. For example, the AGVs could not promptly and accurately receive and execute task instructions issued by the WMS. Real-time AGV status information could not be fed back to the WMS in a timely manner, preventing the WMS from adjusting scheduling strategies based on actual conditions. Summary of the Invention

[0006] The purpose of the present invention is to provide a warehouse scheduling method based on the collaboration of WMS and AGV to solve the problems raised in the above background technology.

[0007] To achieve the above objectives, the present invention provides the following technical solution: a warehouse scheduling method based on the collaboration of WMS and AGV, the method comprising: Build a dynamic storage location data center that includes storage location occupancy status, material turnover rate, and material attribute-related parameters; the material attribute-related parameters include material classification code, storage environment requirements, and in-and-out frequency; Based on the dynamic storage location data center, a dynamic storage location allocation plan is generated through a multi-factor decision algorithm; the multi-factor decision algorithm includes turnover rate weight adjustment, correlation matching calculation and storage location distance coefficient optimization; Based on real-time inbound and outbound instructions, an AGV handling task chain is generated through task decomposition rules; the task decomposition rules include order urgency classification, AGV load balancing, and task node dependencies; Based on the AGV handling task chain, a multi-AGV collaborative path planning network is constructed through a spatiotemporal map; the spatiotemporal map includes the distribution of obstacles in the warehouse area, the real-time location of the AGV, and the path conflict prediction area; The dynamic storage allocation scheme is coupled with the multi-AGV collaborative path planning network input into the scheduling model to generate an AGV cluster scheduling instruction set; Based on the AGV cluster scheduling instruction set, the AGV driving trajectory is adjusted through the anti-collision optimization algorithm to output the final scheduling strategy.

[0008] Preferably, the construction of a dynamic storage location data center includes: Collect historical inbound and outbound data, extract material turnover characteristics and classification association matrix; Divide the storage area into different levels according to the physical parameters of the storage location and mark the temperature and humidity sensitive areas; The material turnover rate characteristics, classification association matrix and warehouse level are spatially mapped to generate a dynamic priority sequence of warehouse locations; The storage location occupancy status is updated based on the dynamic priority sequence to form a dynamic storage location data center.

[0009] Preferably, the generation of a dynamic storage allocation plan by a multi-factor decision algorithm includes: Calculate the product of material turnover rate and storage location distance coefficient as the first allocation weight; Generate a second allocation weight based on the matching degree between the material classification code and the storage environment requirements of the warehouse location; The first allocation weight and the second allocation weight are integrated through a weighted fusion algorithm to generate a dynamic storage score; According to the dynamic storage location scoring sorting results, select the target storage location and update the allocation plan.

[0010] Preferably, generating an AGV handling task chain by using task decomposition rules includes: Parse the order urgency tags in the inbound and outbound instructions and divide the task priority levels; Calculate the available handling capacity based on the AGV's current load status and remaining battery capacity; Generate a task node dependency graph based on the matching relationship between task priority levels and available handling capacity values; Convert the task node dependency graph into an AGV handling task chain and mark the task switching trigger conditions.

[0011] Preferably, the construction of a multi-AGV collaborative path planning network through a spatiotemporal map includes: Collect 3D point cloud data of the storage area and mark the boundaries of fixed obstacles and temporary occupied areas; Generate dynamic traffic weights based on the real-time location of the AGV and mark the predicted path conflict areas; Grid-encode the dynamic traffic weights and the path conflict prediction areas to construct a spatiotemporal map; Generate a multi-AGV collaborative path planning network based on the spatiotemporal map and mark the path priority labels.

[0012] Preferably, the coupling scheduling model includes a path optimization module, and the path optimization module includes: Extract the path conflict prediction area in the multi-AGV collaborative path planning network and calculate the congestion penalty factor; Dynamically adjust path priority labels based on historical execution data of the AGV cluster scheduling instruction set; An iterative optimization algorithm is used to balance the congestion penalty factor and the path priority label to generate an optimized path set. The optimized path set is spatially and temporally aligned with the dynamic storage allocation plan, and the AGV cluster scheduling instruction set is output.

[0013] Preferably, the anti-collision optimization algorithm includes: Predict the position coordinates of the next N time steps based on the AGV's driving trajectory and generate a trajectory conflict detection matrix; Calculate the safety distance offset between each AGV based on the trajectory conflict detection matrix; Adjust the AGV speed or path steering angle through the trajectory replanning algorithm to eliminate the safety distance offset; The adjusted driving trajectory is encoded into the final scheduling strategy, and the spatiotemporal map is updated synchronously.

[0014] Preferably, the method for constructing the classification association matrix includes: Count the co-occurrence frequencies of materials in historical inbound and outbound orders to generate an initial correlation matrix; Normalize the initial correlation matrix according to the material classification coding level; The classification association matrix is updated based on the normalized results, and the priority sequence of the dynamic storage location data center is associated.

[0015] Preferably, the method for generating the dynamic traffic weight includes: Calculate the Euclidean distance between the AGV and the target location based on its real-time position and generate the basic traffic weight; Perform attenuation compensation on the basic traffic weight based on the remaining time of the AGV mission and the remaining battery capacity; Combined with the congestion level of the path conflict prediction area, the access weight is dynamically modified and mapped to the grid code of the space-time map.

[0016] Preferably, the present invention further includes a warehouse scheduling system based on WMS and AGV collaboration, the system comprising: Dynamic location modeling module: used to build a dynamic location data center, which includes location occupancy status, material turnover rate and material attribute related parameters; Storage location allocation module: It is configured to generate dynamic storage location allocation plans based on a dynamic storage location data center through a multi-factor decision algorithm; Task decomposition module: used to generate AGV handling task chains based on real-time inbound and outbound instructions through task decomposition rules; Path planning module: used to build a multi-AGV collaborative path planning network based on the AGV handling task chain through the spatiotemporal map; Coupling scheduling module: The dynamic storage allocation plan and the multi-AGV collaborative path planning network are input into the coupling scheduling model to generate the AGV cluster scheduling instruction set; Anti-collision execution module: Based on the AGV cluster scheduling instruction set, it adjusts the AGV driving trajectory through the anti-collision optimization algorithm and outputs the final scheduling strategy.

[0017] Compared with the prior art, the present invention has the following beneficial effects: The warehouse scheduling method and system based on WMS and AGV collaboration provided by this invention effectively addresses several key issues in traditional warehouse scheduling, resulting in significant benefits in multiple areas. In terms of storage location management, a dynamic storage location data center is constructed to comprehensively integrate information such as storage location occupancy status, material turnover rate, and material attribute-related parameters. Historical inbound and outbound data is used to extract material turnover characteristics and a classification association matrix. Combined with the hierarchical division of storage areas, a precise dynamic priority sequence for storage locations is generated, enabling dynamic storage location management. This allows materials to be rationally allocated to appropriate storage locations based on their characteristics and movement frequency. For example, materials with high turnover rates are preferentially allocated to storage locations near entrances and exits, reducing transportation distances. Materials with strong correlations are stored in adjacent storage locations for simultaneous access, significantly improving storage space utilization and material access efficiency, while avoiding the space waste and difficulty in locating materials associated with traditional fixed storage location allocation methods.

[0018] In the storage location allocation process, a multi-factor decision-making algorithm comprehensively considers turnover rate weight adjustment, correlation matching calculation, and location distance coefficient optimization. By calculating the product of material turnover rate and location distance coefficient, and generating weights based on the matching degree between material classification codes and location storage environment requirements, this weighted fusion generates a dynamic storage location score, which is used to select the target storage location. Compared with traditional single-factor storage location allocation methods, this scientific decision-making approach can more accurately match materials and storage locations, further improving warehouse operation efficiency and reducing operating costs.

[0019] In response to real-time inbound and outbound orders, a rational AGV handling task chain is generated based on task decomposition rules based on order urgency classification, AGV load balancing, and task node dependencies. The order urgency tag is parsed to prioritize tasks. The available handling capacity is calculated based on the AGV's current load and battery level. A task node dependency graph is constructed and converted into a task chain, with task switching trigger conditions noted. This ensures that urgent tasks are prioritized and AGV load is balanced, preventing overwork or idleness of some AGVs due to improper task allocation, thereby improving overall AGV operational efficiency and service life.

[0020] The multi-AGV collaborative path planning network is constructed using spatiotemporal maps. Three-dimensional point cloud data from the warehouse area is collected to annotate obstacles and temporary occupied areas. Dynamic traffic rights are generated based on the real-time positions of the AGVs, and predicted conflict areas are marked. Grid coding is used to construct the map, generate the planning network, and annotate path priority labels. This enables AGVs to avoid obstacles and conflict areas in real time while driving, select the optimal path, effectively reduce path conflicts and congestion, and significantly improve the efficiency and safety of multi-AGV collaborative operations.

[0021] The coupled scheduling model organically combines a dynamic storage allocation plan with a multi-AGV collaborative path planning network. Its path optimization module calculates congestion penalty factors, dynamically adjusts path priority tags, and uses an iterative optimization algorithm to generate an optimized path set. This path set aligns the optimized path with the storage allocation plan in both time and space, outputting a precise set of AGV cluster scheduling instructions. This collaborative scheduling approach ensures smooth integration of warehouse operations, avoids disconnections and conflicts between various links, and improves the accuracy and efficiency of overall scheduling.

[0022] The collision avoidance optimization algorithm predicts the future position of the AGV based on its trajectory, generates a trajectory collision detection matrix, and calculates the safe distance offset. The trajectory replanning algorithm adjusts the speed or path steering angle to eliminate the offset, and then updates the trajectory and space-time map. This algorithm provides a solid guarantee for the safe operation of AGVs in warehouse environments, effectively avoiding collisions, reducing equipment damage and cargo loss, and ensuring stable warehouse operations. BRIEF DESCRIPTION OF THE DRAWINGS

[0023] Figure 1 This is a working principle diagram of the warehouse scheduling method based on WMS and AGV collaboration according to the present invention; Figure 2 Flowchart for building a dynamic location data center; Figure 3 A flow chart for generating a dynamic storage allocation plan; Figure 4 This is the flow chart of the path optimization module in the coupled scheduling model. DETAILED DESCRIPTION

[0024] The following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative efforts are within the scope of protection of the present invention.

[0025] See also Figures 1-4 The present invention provides a warehouse scheduling method and system based on the collaboration of a WMS (Warehouse Management System) and AGVs (Automated Guided Vehicles) to achieve efficient and intelligent warehouse management, improving the overall efficiency and accuracy of warehouse operations. The following describes specific implementations of the present invention.

[0026] The dynamic location data center serves as the foundational data platform for warehouse scheduling. It includes location occupancy status, material turnover rates, and associated material attribute parameters, including material classification codes, storage environment requirements, and inbound and outbound frequency. During its construction, the system first collects historical inbound and outbound data accumulated over a long period of time. This data details information such as the inbound and outbound time and quantity of each material. Through in-depth analysis of this data, a specific data mining algorithm is applied to extract material turnover characteristics and a classification association matrix. Furthermore, the system divides the storage area into hierarchies based on physical parameters such as the floor location and shelf row and column position, and uses temperature and humidity sensors to mark temperature and humidity-sensitive areas. The material turnover characteristics, classification association matrix, and storage area hierarchies are then spatially mapped. Taking these factors into account, a dynamic priority is calculated for each location, generating a dynamic priority sequence. Finally, based on this dynamic priority sequence and combined with real-time material inbound and outbound activity, the location occupancy status is updated, forming the dynamic location data center.

[0027] The multi-factor decision-making algorithm comprehensively considers several key factors, including material turnover rate and storage environment compatibility. First, the product of the material turnover rate and the location distance coefficient is calculated as the first allocation weight. Materials with high material turnover rates tend to be allocated to locations closer to the entrance and exit points, which reduces handling time and improves storage efficiency. The location distance coefficient is calculated based on the actual distance between the location and the entrance and exit points. A second allocation weight is generated based on the compatibility between the material classification code and the location storage environment requirements. For example, temperature-sensitive materials must be allocated to a location that meets their temperature requirements. Next, a weighted fusion algorithm is used to integrate the first and second allocation weights to generate a dynamic storage location score. Based on the ranking of the dynamic storage location scores, the target storage location is selected from high to low scores, and the allocation plan is updated promptly to adapt to dynamic changes in the storage situation.

[0028] After receiving real-time inbound and outbound instructions, the system first parses the order urgency tag in the instruction and divides the tasks into different priority levels. For example, the highest priority is given to expedited order handling tasks. At the same time, the current load status and remaining battery level are obtained through the sensors carried by the AGV itself, and the available handling capacity value is calculated according to a pre-set calculation model. Then, based on the matching relationship between the task priority level and the available handling capacity value, a task node dependency graph is constructed. This graph clearly defines the sequence and dependency relationship between each handling task. Finally, the task node dependency graph is converted into an AGV handling task chain, and the task switching trigger conditions are clearly marked. For example, when the AGV completes a handling task and has sufficient battery power and meets the load requirements, the next task is triggered.

[0029] Spatiotemporal mapping is key to enabling collaborative path planning for multiple AGVs. First, 3D point cloud data of the warehouse area is collected, and advanced 3D scanning technology is used to accurately mark the boundaries of fixed obstacles and temporary occupied areas, such as shelves undergoing repair. Based on the real-time location of the AGV, a specific algorithm generates dynamic rights of passage. Sensors and data analysis technology are used to mark predicted conflict areas. The dynamic rights of passage and predicted conflict areas are grid-encoded to construct an intuitive and accurate spatiotemporal map. Based on the spatiotemporal map, an intelligent path planning algorithm is used to generate a collaborative path planning network for multiple AGVs. Each path is labeled with a path priority tag, and high-priority paths are assigned to AGVs first to improve overall handling efficiency.

[0030] The coupled scheduling model comprehensively considers both storage allocation and path planning. The path optimization module plays a key role. It first extracts predicted conflict areas within the multi-AGV collaborative path planning network and calculates a congestion penalty factor based on the congestion level in that area. Simultaneously, it dynamically adjusts path priority tags based on historical execution data of the AGV cluster scheduling instruction set. An iterative optimization algorithm balances the congestion penalty factor and path priority tags to generate an optimized path set. Finally, the optimized path set is spatially and temporally aligned with the dynamic storage allocation plan to ensure that AGVs arrive at the correct storage location at the correct time for handling operations, thereby outputting the AGV cluster scheduling instruction set.

[0031] The collision avoidance optimization algorithm is a key component in ensuring the safe and efficient operation of AGVs. Based on the AGV's trajectory, a kinematic model is used to predict its position coordinates for the next N time steps, generating a trajectory collision detection matrix. The safety distance offset between each AGV is calculated based on the trajectory collision detection matrix. If the offset exceeds the safe range, a collision risk is indicated. A trajectory replanning algorithm adjusts the AGV's speed or path steering angle to eliminate the safety distance offset. The adjusted trajectory is encoded into the final scheduling strategy, and the spatiotemporal map is simultaneously updated. This allows the system to monitor the AGV's latest position and driving status in real time, enabling continuous and stable warehouse scheduling.

[0032] The technical solution of the present invention is further described in detail below with reference to specific embodiments: Example 1:

[0033] In actual application scenarios, building a dynamic storage location data center is the basis for achieving efficient warehouse scheduling.

[0034] Historical inbound and outbound data is collected. This data covers the inbound and outbound records of all materials in the warehouse over a period of time, including inbound and outbound time, material type, quantity, and corresponding storage location information. Through in-depth analysis of this data, material turnover characteristics are extracted. Material turnover can be calculated as the ratio of the number of materials shipped out to the average inventory quantity over a certain period of time. It reflects the speed of material flow in the warehouse. At the same time, the co-occurrence frequency of materials in historical inbound and outbound orders is counted to generate an initial association matrix. For example, if material A and material B appear simultaneously in multiple orders, the association between them is high. The initial association matrix is normalized according to the material classification coding hierarchy to make the associations between materials at different levels comparable, thereby obtaining a classification association matrix.

[0035] Divide the warehouse into different zones based on the physical parameters of the storage location. For example, divide the warehouse into different areas, such as normal temperature area and cold storage area, based on the warehouse layout and storage conditions. Also, mark temperature and humidity sensitive areas. These areas have specific temperature and humidity requirements for stored materials and require special management.

[0036] By spatially mapping material turnover characteristics, the classification association matrix, and the warehouse hierarchy, we can generate a dynamic priority sequence for warehouse locations. For example, materials with high turnover rates are prioritized for locations near entrances and exits, while materials with strong associations are allocated to adjacent locations as much as possible.

[0037] The occupancy status of storage locations is updated based on a dynamic priority sequence. When new materials are stored or shipped, the occupancy information of the storage locations is updated in a timely manner, forming a dynamic storage location data center. By continuously updating and improving this data center, it provides real-time and accurate data support for subsequent warehouse scheduling.

[0038] Example 2:

[0039] When generating a dynamic storage allocation plan, the multi-factor decision algorithm plays a key role. The product of the material turnover rate and the storage location distance coefficient is calculated as the first allocation weight. Assume that the material turnover rate is , the storage location distance coefficient is , the first allocation weight Material turnover rate As mentioned above, it reflects the flow rate of the material. The larger the value, the faster the material turnover in the warehouse; the location distance coefficient It is determined by the distance between the storage location and the warehouse entrance or exit or key operation area. The closer the distance, The larger the value, the more it can be obtained by calculating the Euclidean distance between the storage location coordinates and the entrance and exit coordinates and normalizing them.

[0040] The second allocation weight is generated based on the matching degree between the material classification code and the location storage environment requirements. Assume that the matching degree is , the second allocation weight . Match It can be calculated according to pre-set rules. For example, if the storage environment required for the material is normal temperature, and a certain storage location is in the normal temperature area, the matching degree between the material and the storage location is 1; if the storage location environment partially matches the material requirements, a corresponding score can be assigned according to the degree of matching, such as 0.5.

[0041] The first allocation weight and the second allocation weight are integrated through the weighted fusion algorithm to generate a dynamic storage score. The weighted coefficients are and , dynamic storage scoring . Weighting coefficient and It can be adjusted according to actual warehouse management needs and experience. For example, if you pay more attention to the impact of material turnover rate on storage space allocation, you can appropriately increase The value of .

[0042] Based on the dynamic storage location score ranking results, select the target storage location and update the allocation plan. Sort the dynamic storage location scores of all storage locations and select the location with the highest score as the target storage location. When new materials need to be stored, recalculate the dynamic storage location score according to the above process and update the allocation plan in a timely manner to ensure the rationality and dynamic adaptability of storage location allocation.

[0043] Example 3:

[0044] After receiving real-time inbound and outbound orders, generating an AGV handling task chain based on task decomposition rules is a key step in achieving efficient warehousing and handling. Order urgency tags in inbound and outbound orders are parsed to prioritize tasks. For example, order urgency is categorized as high, medium, and low. Tasks corresponding to high-urgency orders receive the highest priority and are handled first.

[0045] Calculate the available handling capacity based on the current load status of the AGV and the remaining battery capacity. Assume that the maximum load capacity of the AGV is , the current load is , the remaining battery power is , the battery is fully charged , available handling capacity value This formula takes into account the AGV's load capacity and battery power to ensure that the AGV has sufficient capacity to complete the task and will not be affected by insufficient power during the task.

[0046] Based on the matching relationship between task priority levels and available handling capacity, a task node dependency graph is generated. For example, high-priority tasks are assigned to AGVs with high available handling capacity values first; low-priority tasks are assigned to AGVs with lower remaining available handling capacity values after other high-priority tasks have been assigned. This matching relationship constructs dependency relationships between task nodes, forming a task node dependency graph.

[0047] The task node dependency graph is converted into an AGV handling task chain, and the task switching trigger conditions are annotated. Task switching trigger conditions can include the AGV completing the current task, the battery level falling below a certain threshold, or receiving a higher-priority task. For example, after the AGV completes a handling task, it automatically obtains the next task based on the task node dependency graph. If the battery level falls below 20% during the task execution, a task switch is triggered, and the AGV returns to the charging area for charging. It then resumes the remaining task when the battery is fully charged. By rationally constructing the AGV handling task chain and clarifying the task switching trigger conditions, the AGV's operational efficiency and task execution flexibility can be improved.

[0048] Example 4:

[0049] Building a multi-AGV collaborative path planning network is crucial for achieving efficient, coordinated operation within a warehouse, and spatiotemporal maps are the foundation of this network. Three-dimensional point cloud data of the warehouse area is collected, and using equipment such as LiDAR to obtain 3D spatial information, the boundaries of fixed obstacles and temporary areas are marked. Fixed obstacles include shelves and walls, while temporary areas may include areas where loading and unloading operations are ongoing or where goods are temporarily stored.

[0050] Generate dynamic traffic weights based on the real-time location of the AGV. Assume that the real-time location of the AGV is , the target storage location is , calculate the Euclidean distance between it and the target location according to the Euclidean distance formula , generate basic traffic weight Then, the basic traffic weight is attenuated and compensated based on the remaining time of the AGV mission and the remaining battery capacity. Assume that the remaining time of the mission is , the remaining battery power is , the battery is fully charged , attenuation compensation coefficient ( is the total mission time), the adjusted traffic weight Finally, the access weight is dynamically adjusted based on the congestion level of the predicted conflict area. If the congestion level in the predicted conflict area is high, the access weight is reduced; if the congestion level is low, the access weight is appropriately increased. The adjusted access weight is mapped to the grid code of the spatiotemporal map.

[0051] At the same time, path conflict prediction areas are marked based on the AGV's real-time location and travel direction. When multiple AGVs are likely to reach the same location or intersect paths at a given moment, this is considered a path conflict prediction area. Dynamic traffic weights and path conflict prediction areas are grid-encoded to construct a spatiotemporal map. Each grid in the spatiotemporal map contains information such as the traffic weight for that area, the presence of obstacles, and path conflict predictions.

[0052] A multi-AGV collaborative path planning network is generated based on the spatiotemporal map and path priority labels are annotated. Path priority labels are determined based on factors such as task priority and access weight. For example, paths for high-priority tasks are assigned a higher path priority label; paths with high access weights are also assigned a higher path priority label. By constructing a multi-AGV collaborative path planning network and annotating path priority labels, reasonable driving path planning is provided for AGVs, avoiding path conflicts and improving the efficiency of multi-AGV collaborative operations.

[0053] Example 5:

[0054] The path optimization module in the coupled scheduling model is crucial for generating an efficient AGV cluster scheduling instruction set. First, the path conflict prediction area in the multi-AGV collaborative path planning network is extracted and the congestion penalty factor is calculated. Let the number of AGVs expected to conflict in the path conflict prediction area be , the area of this region is , congestion penalty factor The more AGVs that conflict, the smaller the area, and the larger the congestion penalty factor, the higher the congestion level in the area.

[0055] Based on historical execution data from the AGV cluster's scheduling instructions, path priority tags are dynamically adjusted. For example, if a path frequently experiences congestion or delays during historical execution, its path priority tag is lowered; if a path demonstrates high execution efficiency, its path priority tag is increased. By analyzing and learning from historical data, path priority tags are continuously optimized to better reflect actual operational conditions.

[0056] The congestion penalty factor and the path priority label are balanced through an iterative optimization algorithm to generate an optimized path set. Common iterative optimization algorithms include genetic algorithms and simulated annealing algorithms. Taking the genetic algorithm as an example, the path is regarded as an individual, and the population is continuously evolved through operations such as selection, crossover, and mutation, so that the optimized path has the best overall performance while considering the congestion penalty factor and the path priority label. In each iteration, the fitness value of each individual is calculated. The fitness value can be determined comprehensively based on the congestion penalty factor and the path priority label, such as the fitness value ( and is the weight coefficient, and ). Through multiple iterations, an optimized path set is finally generated.

[0057] The optimized path set is spatially and temporally aligned with the dynamic storage allocation plan to generate the AGV cluster scheduling instruction set. This spatial and temporal alignment involves matching the optimized paths with the material storage locations and entry and exit times, ensuring that the AGVs can execute their handling tasks in a reasonable order and on a timely basis. For example, the dynamic storage allocation plan determines the material storage location. Combined with the optimized path set, a specific driving path and task execution time are generated for each AGV, forming the AGV cluster scheduling instruction set that guides the AGV's actual operations.

[0058] Example 6:

[0059] Anti-collision optimization algorithm is a key technology to ensure the safe operation of AGV in the warehouse. The position coordinates of the time step. Let the current position coordinates of AGV be , the driving speed is , the driving direction angle is , at time step ( ) can be obtained by the formula Calculated ( The time interval for each time step). Based on these predicted position coordinates, a trajectory conflict detection matrix is generated. If the distance between the predicted position coordinates of two AGVs at a certain moment in the future is less than the safety distance threshold , the corresponding position in the trajectory conflict detection matrix is marked as 1, indicating that there is a possibility of conflict; otherwise it is marked as 0.

[0060] The safe distance offset between each AGV is calculated based on the trajectory conflict detection matrix. Assume that the predicted position coordinates of two AGVs with possible conflicts are and , safety distance offset .

[0061] Trajectory replanning algorithms adjust the AGV's speed or path steering angle to eliminate safety distance offsets. For example, if a potential collision between two AGVs is detected, the speed of one AGV can be reduced or its path steering angle adjusted to avoid the conflicting area. Trajectory replanning algorithms can utilize modified versions of classic path planning algorithms, such as the Dijkstra algorithm, to replan the AGV's trajectory while taking into account safety distance offsets and other constraints.

[0062] The adjusted driving trajectory is encoded into the final scheduling strategy, and the spatiotemporal map is updated simultaneously. The adjusted driving trajectory information is converted into a command format that the AGV can recognize and execute, and output as the final scheduling strategy. Simultaneously, the spatiotemporal map is updated with information such as the AGV's location and predicted path conflict areas, so that subsequent path planning and scheduling decisions can be made based on the latest status, ensuring safe and stable operation of warehousing operations.

[0063] It should be noted that, in this document, relational terms such as first and second, etc., are used only to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any actual relationship or order between these entities or operations. Moreover, the terms "comprises," "includes," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that includes a list of elements includes not only those elements but also other elements not explicitly listed, or elements inherent to such process, method, article, or apparatus.

[0064] While embodiments of the present invention have been shown and described, it will be appreciated by those skilled in the art that various changes, modifications, substitutions, and variations may be made to these embodiments without departing from the principles and spirit of the invention, and that the scope of the invention is defined by the appended claims and their equivalents.

Claims

1. A warehouse scheduling method based on WMS and AGV collaboration, characterized in that: include: Build a dynamic storage location data center that includes storage location occupancy status, material turnover rate, and material attribute-related parameters; the material attribute-related parameters include material classification code, storage environment requirements, and in-and-out frequency; Based on the dynamic storage location data center, a dynamic storage location allocation plan is generated through a multi-factor decision algorithm; the multi-factor decision algorithm includes turnover rate weight adjustment, correlation matching calculation and storage location distance coefficient optimization; Based on real-time inbound and outbound instructions, an AGV handling task chain is generated through task decomposition rules; the task decomposition rules include order urgency classification, AGV load balancing, and task node dependencies; Based on the AGV handling task chain, a multi-AGV collaborative path planning network is constructed through a spatiotemporal map; the spatiotemporal map includes the distribution of obstacles in the warehouse area, the real-time location of the AGV, and the path conflict prediction area; The dynamic storage allocation scheme is coupled with the multi-AGV collaborative path planning network input into the scheduling model to generate an AGV cluster scheduling instruction set; Based on the AGV cluster scheduling instruction set, the AGV driving trajectory is adjusted through the anti-collision optimization algorithm to output the final scheduling strategy.

2. A warehouse scheduling method based on WMS and AGV collaboration according to claim 1, characterized in that: The construction of a dynamic storage location data center includes: Collect historical inbound and outbound data, extract material turnover characteristics and classification association matrix; Divide the storage area into different levels according to the physical parameters of the storage location and mark the temperature and humidity sensitive areas; The material turnover rate characteristics, classification association matrix and warehouse level are spatially mapped to generate a dynamic priority sequence of warehouse locations; The storage location occupancy status is updated based on the dynamic priority sequence to form a dynamic storage location data center.

3. The warehouse scheduling method based on WMS and AGV collaboration according to claim 1 is characterized in that: The method of generating a dynamic storage allocation plan by a multi-factor decision algorithm includes: Calculate the product of material turnover rate and storage location distance coefficient as the first allocation weight; Generate a second allocation weight based on the matching degree between the material classification code and the storage environment requirements of the warehouse location; The first allocation weight and the second allocation weight are integrated through a weighted fusion algorithm to generate a dynamic storage score; According to the dynamic storage location scoring sorting results, select the target storage location and update the allocation plan.

4. The warehouse scheduling method based on WMS and AGV collaboration according to claim 1 is characterized in that: The generation of the AGV handling task chain by the task decomposition rule includes: Parse the order urgency tags in the inbound and outbound instructions and divide the task priority levels; Calculate the available handling capacity based on the AGV's current load status and remaining battery capacity; Generate a task node dependency graph based on the matching relationship between task priority levels and available handling capacity values; Convert the task node dependency graph into an AGV handling task chain and mark the task switching trigger conditions.

5. The warehouse scheduling method based on WMS and AGV collaboration according to claim 1 is characterized in that: The multi-AGV collaborative path planning network is constructed through the spatiotemporal map, including: Collect 3D point cloud data of the storage area and mark the boundaries of fixed obstacles and temporary occupied areas; Generate dynamic traffic weights based on the real-time location of the AGV and mark the predicted path conflict areas; Grid-encode the dynamic traffic weights and the path conflict prediction areas to construct a spatiotemporal map; Generate a multi-AGV collaborative path planning network based on the spatiotemporal map and mark the path priority labels.

6. The warehouse scheduling method based on WMS and AGV collaboration according to claim 1 is characterized in that: The coupled scheduling model includes a path optimization module, which includes: Extract the path conflict prediction area in the multi-AGV collaborative path planning network and calculate the congestion penalty factor; Dynamically adjust path priority labels based on historical execution data of the AGV cluster scheduling instruction set; An iterative optimization algorithm is used to balance the congestion penalty factor and the path priority label to generate an optimized path set. The optimized path set is spatially and temporally aligned with the dynamic storage allocation plan, and the AGV cluster scheduling instruction set is output.

7. The warehouse scheduling method based on WMS and AGV collaboration according to claim 1 is characterized in that: The anti-collision optimization algorithm includes: Predict the position coordinates of the next N time steps based on the AGV's driving trajectory and generate a trajectory conflict detection matrix; Calculate the safety distance offset between each AGV based on the trajectory conflict detection matrix; Adjust the AGV speed or path steering angle through the trajectory replanning algorithm to eliminate the safety distance offset; The adjusted driving trajectory is encoded into the final scheduling strategy, and the spatiotemporal map is updated synchronously.

8. The warehouse scheduling method based on WMS and AGV collaboration according to claim 2 is characterized in that: The method for constructing the classification association matrix includes: Count the co-occurrence frequencies of materials in historical inbound and outbound orders to generate an initial correlation matrix; Normalize the initial correlation matrix according to the material classification coding level; The classification association matrix is updated based on the normalized results, and the priority sequence of the dynamic storage location data center is associated.

9. The warehouse scheduling method based on WMS and AGV collaboration according to claim 5 is characterized in that: The method for generating the dynamic traffic weight includes: Calculate the Euclidean distance between the AGV and the target location based on its real-time position and generate the basic traffic weight; Perform attenuation compensation on the basic traffic weight based on the remaining time of the AGV mission and the remaining battery capacity; Combined with the congestion level of the path conflict prediction area, the access weight is dynamically modified and mapped to the grid code of the space-time map.

10. A warehouse scheduling system based on WMS and AGV collaboration, characterized in that: include: Dynamic location modeling module: used to build a dynamic location data center, which includes location occupancy status, material turnover rate and material attribute related parameters; Storage location allocation module: It is configured to generate dynamic storage location allocation plans based on a dynamic storage location data center through a multi-factor decision algorithm; Task decomposition module: used to generate AGV handling task chains based on real-time inbound and outbound instructions through task decomposition rules; Path planning module: used to build a multi-AGV collaborative path planning network based on the AGV handling task chain through the spatiotemporal map; Coupling scheduling module: The dynamic storage allocation plan and the multi-AGV collaborative path planning network are input into the coupling scheduling model to generate the AGV cluster scheduling instruction set; Anti-collision execution module: Based on the AGV cluster scheduling instruction set, it adjusts the AGV driving trajectory through the anti-collision optimization algorithm and outputs the final scheduling strategy.

Images