An intelligent warehouse management method and system capable of improving warehouse management scheduling efficiency

Abstract

The application provides an intelligent warehouse management method and system capable of improving warehouse management scheduling efficiency, relates to the technical field of warehouse scheduling, acquires warehouse information of a warehouse, divides the warehouse into a warehouse-in buffer zone, a dynamic picking zone and a warehouse-out temporary storage zone based on a spatial gridding method and in combination with a accessibility analysis method, acquires historical warehouse-in and warehouse-out information of the warehouse, adjusts the area boundaries of the warehouse-in buffer zone, the dynamic picking zone and the warehouse-out temporary storage zone by using a long short-term memory network, and performs task allocation and path collaborative scheduling on a goods carrier based on a warehouse-in and warehouse-out coordination consistency control algorithm, constructs a five-dimensional twin model of the warehouse by using digital twin technology, continuously monitors, adjusts and optimizes the warehouse-in and warehouse-out operation processes in the warehouse by using the five-dimensional twin model of the warehouse, and realizes collaborative scheduling of the warehouse-in and warehouse-out. The application can effectively avoid task conflicts and resource waste, improve carrying efficiency and reduce operating costs.

Description

Technical Field

[0001] This invention relates to the field of warehouse scheduling technology, and more specifically, to an intelligent warehouse management method and system that can improve the efficiency of warehouse management and scheduling. Background Technology

[0002] Warehouse management has always been a crucial part of the logistics system, primarily responsible for providing basic inventory information, including a wealth of logistics data such as material receipt, shipment, and material management. However, compared to the overall development of the logistics industry, the development of warehouse management has been relatively slow, with its level of informatization and specialization lagging behind.

[0003] In traditional logistics warehousing management, areas are often roughly divided based on experience, failing to fully consider the actual spatial layout of the warehouse, the storage characteristics of the goods, and the needs of the operational processes. This leads to low warehouse space utilization, with some areas potentially overcrowded while others remain idle and wasted. Furthermore, the allocation of tasks and route planning for cargo handling vehicles typically relies on manual methods or simple scheduling rules, making it difficult to dynamically adjust based on real-time inbound and outbound demand and the status of the vehicles. This can easily result in unreasonable task allocation, with some vehicles overloaded while others remain idle, leading to resource waste.

[0004] No effective solutions have yet been proposed to address the problems in the relevant technologies. Summary of the Invention

[0005] In view of this, the present invention provides an intelligent warehouse management method and system that can improve the efficiency of warehouse management and scheduling, so as to solve the above-mentioned problems.

[0006] To solve the above problems, the specific technical solution adopted by the present invention is as follows:

[0007] According to one aspect of the present invention, an intelligent warehouse management method is provided that can improve the efficiency of warehouse management and scheduling, comprising the following steps:

[0008] S1. Obtain the storage space area of ​​the warehouse, grid the storage space area, and determine the location information of each grid cell based on the coordinate transformation method; analyze the storage capacity, cargo turnover characteristics and related operation requirements of each grid cell, and configure functional attribute labels for each grid cell; based on the location information of each grid cell, use accessibility analysis to simulate the operation process in the warehouse, and perform accessibility analysis on each grid cell to obtain accessibility analysis results; combine the functional attribute labels and accessibility analysis results to cluster the grid cells into inbound buffer zone, dynamic picking zone, and outbound temporary storage zone;

[0009] S2. Obtain historical inbound and outbound information from the warehouse, use a long short-term memory network to adjust the boundaries of the inbound buffer zone, dynamic picking zone, and outbound temporary storage zone, and use an inbound and outbound coordination and consistency control algorithm to allocate tasks and coordinate the path scheduling of the cargo handling vehicles.

[0010] Preferably, the analysis of the storage capacity, cargo turnover characteristics, and related operational requirements of each grid cell, and the configuration of functional attribute labels for each grid cell, includes the following steps:

[0011] Collect the structural information and static attribute data of each grid cell, and combine them with the historical inbound and outbound records of each grid cell to analyze the historical cargo storage type, storage frequency and turnover cycle of each grid cell;

[0012] Based on structural information, static attribute data, and historical cargo storage conditions, the fuzzy comprehensive evaluation method is used to assess the storage capacity of each grid unit, and cargo turnover characteristics are summarized based on cargo turnover cycle and storage frequency.

[0013] Based on the assessment results, the storage capacity, cargo turnover characteristics and operational requirements of each grid unit are correlated and matched to determine the functional type of each grid unit;

[0014] Configure a functional attribute label for each grid cell based on its functional type.

[0015] Preferably, the step of simulating the warehouse workflow using accessibility analysis based on the location information of each grid cell, and performing accessibility analysis on each grid cell to obtain the accessibility analysis results includes the following steps:

[0016] Based on the structural characteristics of the warehouse, a simulation model of the warehouse is established using the finite element analysis method, and the warehouse entrance is taken as the starting point of the path.

[0017] Starting from the path origin, the operation process within the warehouse is simulated, and a path search algorithm is used to calculate the optimal path from the path origin to each grid cell and its corresponding path cost, thus obtaining the path calculation results.

[0018] Based on the path calculation results, and using the coverage analysis method, accessibility is evaluated for each grid cell to obtain accessibility analysis results.

[0019] Preferably, the accessibility assessment of each grid cell based on the path calculation results and using the coverage analysis method to obtain the accessibility analysis results includes the following steps:

[0020] Centered on the inlet, the evaluation of each grid cell is carried outward layer by layer, and the path calculation results of each grid cell in each layer are collected.

[0021] Based on the distribution of obstacles in the warehouse and the information of the cargo handling vehicles, the occlusion angle corresponding to each grid cell is calculated, and the path quality of each grid cell is evaluated in combination with the path cost.

[0022] Based on the different work processes of each grid cell, the coverage of each grid cell is analyzed to determine the coverage quality of each grid cell;

[0023] Based on the path quality and coverage quality of each grid cell, accessibility is assessed for each grid cell, and accessibility analysis results are generated.

[0024] Preferably, the formula for calculating the shading angle corresponding to each grid cell based on the obstacle distribution and cargo handling vehicle information within the warehouse is as follows:

[0025] ;

[0026] In the formula, i obs Indicates the shading angle. h obs Indicates the height of the obstacle's vertex. h AGV This indicates the height of the top of the cargo handling vehicle. d ij Represents grid cells i To the observation point j Horizontal distance l Indicates the dynamic obstacle sensitivity coefficient. d dynamic This indicates the amount of dynamic obstacle correction.

[0027] Preferably, the step of combining functional attribute tags and accessibility analysis results to cluster grid cells into inbound buffer zones, dynamic picking zones, and outbound temporary storage zones includes the following steps:

[0028] Based on the functional attribute labels and accessibility analysis results, a clustering index system is constructed;

[0029] The inbound buffer zone, dynamic picking zone, and outbound temporary storage zone are selected as target categories, and the number of clusters is configured accordingly.

[0030] The number of cluster centers is selected randomly from all grid cells.

[0031] Calculate the distance from each grid cell to the initial cluster center, and assign the grid cells to the corresponding target categories according to the nearest distance principle;

[0032] Recalculate the cluster center for each target category, and repeat the process of calculating distance and assigning categories until the cluster centers no longer change, thus obtaining the final clustering result.

[0033] Preferably, the steps of acquiring historical warehouse inbound and outbound information, adjusting the boundaries of the inbound buffer zone, dynamic picking zone, and outbound temporary storage zone using a long short-term memory network, and allocating tasks and coordinating path scheduling for cargo handling vehicles based on an inbound / outbound coordination and consistency control algorithm include the following steps:

[0034] S21. Using a long short-term memory network, construct an inbound / outbound prediction model, and train the inbound / outbound prediction model using historical warehouse inbound and outbound information.

[0035] S22. Based on the trained inbound and outbound prediction model, predict the inbound and outbound quantities of the warehouse, and calculate the demand deviation based on the predicted inbound and outbound quantities.

[0036] S23. Based on demand deviation and combined with preset deviation thresholds, adjust the distinction boundaries of the inbound buffer zone, dynamic picking zone and outbound temporary storage zone respectively.

[0037] S24. Based on the predicted inbound and outbound volumes, the inbound and outbound coordination and consistency control algorithm is used to allocate tasks and coordinate the path scheduling of the cargo handling vehicles.

[0038] Preferably, the step of allocating tasks and coordinating path scheduling for cargo handling vehicles based on predicted inbound and outbound volumes using an inbound / outbound coordination and consistency control algorithm includes the following steps:

[0039] S241. Based on the area information of the inbound buffer zone, dynamic picking zone, and outbound temporary storage zone, as well as the configuration information of the cargo handling vehicle, construct the warehouse-level intelligent agent and the cargo handling vehicle-level intelligent agent.

[0040] S242. Based on the predicted inbound and outbound volumes, determine the target task volume for the inbound buffer zone, dynamic picking zone, and outbound temporary storage zone, respectively.

[0041] S243. Based on the target task volume of the inbound buffer, dynamic picking area and outbound temporary storage area, the warehouse-level intelligent agent coordinates the task requirements among the inbound buffer, dynamic picking area and outbound temporary storage area through a consistency control algorithm.

[0042] S244. The warehouse-level agent distributes the coordinated task requirements to the transport vehicle-level agents in the inbound buffer, dynamic picking area, and outbound temporary storage area, and optimizes the task allocation and path planning of each transport vehicle through a consistency control algorithm.

[0043] Preferably, based on the target task volume of the inbound buffer, dynamic picking area, and outbound temporary storage area, the warehouse-level intelligent agent coordinates the task requirements among the inbound buffer, dynamic picking area, and outbound temporary storage area through a consistency control algorithm, including the following steps:

[0044] S2431. The target task volume of the inbound buffer, dynamic picking area and outbound temporary storage area is modeled as an equivalent task load.

[0045] S2432. Construct a communication topology diagram between the inbound buffer, dynamic picking area and outbound temporary storage area, and construct an adjacency matrix based on the inter-region task migration and resource sharing relationship;

[0046] S2433. Using the discrete consistency iterative formula, the task demand growth rate of the warehouse-level intelligent agent is coordinated to achieve consistency through the warehouse-level intelligent agent.

[0047] Preferably, the process of acquiring historical warehouse inbound and outbound information, adjusting the boundaries of the inbound buffer zone, dynamic picking zone, and outbound temporary storage zone using a long short-term memory network, and allocating tasks and coordinating path scheduling for the cargo handling vehicles based on an inbound / outbound coordination and consistency control algorithm includes:

[0048] A five-dimensional twin model of the warehouse is constructed using digital twin technology. This model is then used to continuously monitor, adjust, and optimize the inbound and outbound operations within the warehouse, thereby achieving coordinated scheduling of inbound and outbound processes.

[0049] According to one aspect of the present invention, an intelligent warehouse management system is provided that can improve the efficiency of warehouse management and scheduling. The system includes:

[0050] The area division module is used to obtain warehouse storage information. Based on the spatial gridding method and combined with the accessibility analysis method, the warehouse is divided into an inbound buffer zone, a dynamic picking zone, and an outbound temporary storage zone.

[0051] The collaborative scheduling module is used to obtain historical warehouse inbound and outbound information, use a long short-term memory network to adjust the regional boundaries of the inbound buffer zone, dynamic picking zone and outbound temporary storage zone, and perform task allocation and path collaborative scheduling for cargo handling vehicles based on the inbound and outbound coordination and consistency control algorithm.

[0052] The beneficial effects of this invention are as follows:

[0053] 1. This invention divides the warehouse into an inbound buffer zone, a dynamic picking zone, and an outbound temporary storage zone using a spatial gridding method combined with accessibility analysis. This makes the zone division more scientific and reasonable, which helps to improve warehouse space utilization. Based on the inbound and outbound coordination and consistency control algorithm, the invention allocates tasks and coordinates the paths of the cargo handling vehicles, which can effectively avoid task conflicts and resource waste, improve handling efficiency, and reduce operating costs. The invention also constructs a five-dimensional twin model, which can not only continuously monitor the warehouse inbound and outbound operation process and promptly detect and warn of problems, but also adjust and optimize the operation process through simulation and evaluation, thereby achieving coordinated scheduling of inbound and outbound operations.

[0054] 2. This invention, by collecting multi-dimensional data and using scientific methods such as fuzzy comprehensive evaluation, can accurately grasp the characteristics of each grid unit, providing a strong basis for the rational division of areas. It uses accessibility analysis to simulate the operation process and analyze the accessibility of grid units. By establishing simulation models, calculating the optimal path and cost, and combining coverage analysis, it comprehensively evaluates the path quality and coverage quality of grid units, ensuring that the area division can meet the accessibility requirements of efficient operation.

[0055] 3. In terms of task allocation and path coordination scheduling for cargo handling vehicles, this invention constructs warehouse-level and cargo handling vehicle-level intelligent agents. Based on the predicted task volume, the target task volume for each area is determined. The warehouse-level intelligent agent coordinates the task requirements between areas using a consistency control algorithm, and distributes the coordinated tasks to the cargo handling vehicle-level intelligent agent to optimize task allocation and path planning. This improves the operational efficiency of cargo handling vehicles, reduces operating costs, enhances the coordination and flexibility of the overall warehouse operation, and helps to improve the intelligence level and overall efficiency of warehouse management. Attached Figure Description

[0056] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the embodiments will be briefly described below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort. In the drawings:

[0057] Figure 1 This is a flowchart of an intelligent warehouse management method that can improve warehouse management and scheduling efficiency according to an embodiment of the present invention;

[0058] Figure 2 This is a schematic diagram of an intelligent warehouse management system that can improve warehouse management and scheduling efficiency according to an embodiment of the present invention.

[0059] In the picture:

[0060] 1. Region division module; 2. Collaborative scheduling module; 3. Job adjustment module. Detailed Implementation

[0061] To enable those skilled in the art to better understand the technical solutions in this application, the technical solutions in the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only a part of the embodiments of this application, and not all of the embodiments. Based on the embodiments in this application, all other embodiments obtained by those skilled in the art without creative effort should fall within the scope of protection of this application.

[0062] According to embodiments of the present invention, an intelligent warehouse management method and system are provided that can improve the efficiency of warehouse management and scheduling.

[0063] The present invention will now be further described in conjunction with the accompanying drawings and specific embodiments, such as... Figure 1 As shown, according to an embodiment of the present invention, an intelligent warehouse management method is provided that can improve the efficiency of warehouse management and scheduling, including the following steps:

[0064] S1. Obtain warehouse storage information, and based on the spatial gridding method and combined with accessibility analysis, divide the warehouse into an inbound buffer zone, a dynamic picking zone, and an outbound temporary storage zone.

[0065] As a preferred embodiment, the acquisition of warehouse storage information, based on the spatial gridding method and combined with accessibility analysis, divides the warehouse into an inbound buffer zone, a dynamic picking zone, and an outbound temporary storage zone, including the following steps:

[0066] S11. Obtain the storage space area of ​​the warehouse, perform grid processing on the storage space area to obtain several grid cells, and determine the location information of each grid cell based on the coordinate transformation method;

[0067] It should be noted that the warehousing information includes the storage space area of ​​the warehouse. When obtaining the storage space area of ​​the warehouse, it can be collected through the warehouse's architectural design drawings, actual measurements, or spatial data in the warehouse management system (WMS). Then, the obtained warehouse storage space area is divided into several uniform or non-uniform grid units according to certain rules.

[0068] For each grid cell, a two-dimensional or three-dimensional coordinate system is used to represent its position. Then, a coordinate transformation method is used to assign a unique position identifier to each grid cell. For example, in a two-dimensional planar coordinate system, with a corner of the warehouse as the origin, the coordinate values ​​of each grid cell in the horizontal and vertical directions are determined.

[0069] S12. Analyze the storage capacity, cargo turnover characteristics and related operational requirements of each grid cell, and configure functional attribute labels for each grid cell.

[0070] As a preferred embodiment, the analysis of the storage capacity, cargo turnover characteristics, and related operational requirements of each grid cell, and the configuration of functional attribute labels for each grid cell, includes the following steps:

[0071] S121. Collect the structural information and static attribute data of each grid cell, and combine them with the historical inbound and outbound records of each grid cell to analyze the historical cargo storage type, storage frequency and turnover cycle of each grid cell.

[0072] Specifically, the structural information of each grid cell includes the grid cell's dimensions (length, width, height), shape, load-bearing capacity, etc., while the static attribute data of each grid cell includes the grid cell's number, its location, and the condition of surrounding facilities (such as shelves, aisles, etc.).

[0073] S122. Based on structural information, static attribute data, and historical cargo storage conditions, the fuzzy comprehensive evaluation method is used to assess the storage capacity of each grid unit, and cargo turnover characteristics are summarized according to cargo turnover cycle and storage frequency.

[0074] It should be noted that the fuzzy comprehensive evaluation method is a comprehensive evaluation method based on fuzzy mathematics. When evaluating the storage capacity of a grid cell, multiple factors need to be considered, such as the size of the grid cell, its load-bearing capacity, and the surrounding environment. Different weights and evaluation levels are assigned to each factor. Then, by establishing a fuzzy evaluation matrix, the storage capacity of each grid cell is comprehensively evaluated to obtain a relatively accurate evaluation result, such as dividing the storage capacity into different levels such as high, medium, and low.

[0075] Furthermore, based on data on cargo turnover cycle and storage frequency, the cargo turnover characteristics of each grid unit are summarized. For example, grid units with short turnover cycles and high storage frequency are suitable for storing high-frequency turnover goods, while grid units with long turnover cycles and low storage frequency are more suitable for storing long-term storage goods. By summarizing cargo turnover characteristics, we can better understand the advantages and characteristics of each grid unit in terms of cargo turnover.

[0076] S123. Based on the evaluation results, the storage capacity, cargo turnover characteristics and operational requirements of each grid unit are correlated and matched to determine the functional type of each grid unit.

[0077] Specifically, operational requirements include storage needs for different types of goods, inbound / outbound frequency requirements, and operational limitations of handling equipment. For goods requiring high-frequency inbound / outbound movement and with relatively light weight, grid units with moderate storage capacity and high-frequency goods turnover should be matched. For large, heavy goods requiring long-term storage, grid units with high storage capacity and low-frequency goods turnover should be matched. Through correlation matching, the functional type of each grid unit is determined, such as high capacity, high flexibility in goods turnover, and ease of quick retrieval by operators.

[0078] S124. Configure a functional attribute label for each grid cell based on its functional type.

[0079] S13. Based on the location information of each grid cell, use accessibility analysis to simulate the operation process in the warehouse, and perform accessibility analysis on each grid cell to obtain the accessibility analysis results.

[0080] In a preferred embodiment, the step of simulating the warehouse workflow using accessibility analysis based on the location information of each grid cell, and performing accessibility analysis on each grid cell to obtain the accessibility analysis results includes the following steps:

[0081] S131. Based on the structural characteristics of the warehouse, a simulation model of the warehouse is established using the finite element analysis method, and the warehouse entrance is taken as the starting point of the path.

[0082] S132. Starting from the path origin, simulate the work process within the warehouse and use the path search algorithm to calculate the optimal path from the path origin to each grid cell and its corresponding path cost, and obtain the path calculation results.

[0083] It should be noted that, based on the simulation model, starting from the warehouse entrance, the simulation simulates the operations of goods entering, handling, and storing, following the actual operating rules and processes of the warehouse.

[0084] Among them, path search algorithms (such as A* algorithm, Dijkstra algorithm, etc.) are used to find the optimal path from the path origin (entry point) to each grid cell in the simulation model. The optimal path is the path that meets certain conditions (such as shortest distance, minimum time, etc.); at the same time, the path cost corresponding to each path is calculated, which includes multiple factors such as distance, time, energy consumption, etc.

[0085] S133. Based on the path calculation results and using the coverage analysis method, the accessibility of each grid cell is evaluated to obtain the accessibility analysis results.

[0086] In a preferred embodiment, the accessibility assessment of each grid cell based on the path calculation results and using the coverage analysis method to obtain the accessibility analysis results includes the following steps:

[0087] S1331. Taking the inlet as the center, evaluate each grid cell by diffusing outward layer by layer, and collect the path calculation results of each grid cell in each layer.

[0088] It should be noted that expanding outwards from the warehouse entrance layer by layer means dividing the warehouse space into multiple levels, each representing a different grid area. The expansion of each level collects path calculation results, providing foundational data for subsequent accessibility analysis.

[0089] S1332. Based on the distribution of obstacles in the warehouse and the information of the cargo handling vehicles, calculate the occlusion angle corresponding to each grid cell, and evaluate the path quality of each grid cell in combination with the path cost.

[0090] Specifically, the occlusion angle refers to the angle at which the view of a grid cell is restricted due to obstacles or the presence of transport vehicles within the warehouse. Calculating the occlusion angle helps analyze the visibility and accessibility of a path. The size of the occlusion angle directly affects the quality of the path, as a larger occlusion angle may indicate that the path is blocked by obstacles or other factors.

[0091] Among them, path cost reflects the economy and efficiency of a path, while occlusion angle reflects the accessibility of the path. For example, a grid cell with low path cost but a large occlusion angle may have a short path but be difficult to pass through; a grid cell with high path cost but a small occlusion angle may have smooth passage but poor economy. By taking both factors into account, the path quality of each grid cell can be evaluated more comprehensively.

[0092] In a preferred embodiment, the formula for calculating the shading angle corresponding to each grid cell based on the obstacle distribution and cargo handling vehicle information within the warehouse is as follows:

[0093] ;

[0094] In the formula, i obs Indicates the shading angle. h obs Indicates the height of the obstacle's vertex. h AGV This indicates the height of the top of the cargo handling vehicle. d ij Represents grid cells i To the observation point j Horizontal distance l The dynamic obstacle sensitivity coefficient reflects the impact of dynamic changes in obstacles on the occlusion angle. d dynamic This represents the dynamic obstacle correction amount, taking into account the impact of the dynamic environment within the warehouse, such as other moving cargo handling vehicles or personnel.

[0095] S1333. Based on the different work processes of each grid cell, analyze the coverage of each grid cell and determine the coverage quality of each grid cell;

[0096] Coverage refers to whether the paths and operational functions of a grid cell can cover other areas or work points in the warehouse. Based on the coverage analysis results, the coverage quality of each grid cell is determined. Coverage quality can be divided into different levels, such as high, medium, and low. Grid cells with high coverage quality have a high usage frequency and importance in the workflow, and their accessibility needs to be guaranteed. Grid cells with low coverage quality may have accessibility requirements appropriately reduced in some cases.

[0097] S1334. Based on the path quality and coverage quality of each grid cell, perform accessibility assessment on each grid cell and generate accessibility analysis results.

[0098] It should be noted that both path quality and coverage quality are important factors affecting accessibility. For example, grid cells with high path quality and high coverage quality have good accessibility, while grid cells with low path quality or low coverage quality have poor accessibility. A weighted average method can be used to integrate path quality and coverage quality to obtain an accessibility assessment value for each grid cell.

[0099] S14. Combining the functional attribute labels and accessibility analysis results, the grid cells are clustered into inbound buffer zone, dynamic picking zone, and outbound temporary storage zone.

[0100] As a preferred embodiment, the step of clustering grid cells into inbound buffer, dynamic picking area, and outbound temporary storage area by combining functional attribute tags and accessibility analysis results includes the following steps:

[0101] S141. Based on the functional attribute labels and accessibility analysis results, construct a clustering index system;

[0102] Specifically, a clustering index system is constructed by selecting indicators such as carrying capacity and cargo turnover frequency from the functional attribute tags, as well as path length and obstacle impact from the accessibility analysis results.

[0103] S142. Designate the inbound buffer zone, dynamic picking zone, and outbound temporary storage zone as target categories, and configure the number of clusters.

[0104] Specifically, since the inbound buffer, dynamic picking area, and outbound temporary storage area are taken as target categories, the number of clusters is set to 3.

[0105] S143. Select the number of cluster cells as the initial cluster centers from all grid cells using a random selection method;

[0106] S144. Calculate the distance from each grid cell to the initial cluster center, and assign the grid cells to the corresponding target categories according to the nearest distance principle;

[0107] S145. Recalculate the cluster center for each target category, and repeat the process of calculating distance and assigning categories until the cluster centers no longer change, thus obtaining the final clustering result.

[0108] Specifically, by collecting multi-dimensional data and using scientific methods such as fuzzy comprehensive evaluation, we can accurately grasp the characteristics of each grid unit, providing a strong basis for the rational division of areas. By using accessibility analysis to simulate the operation process and analyze the accessibility of grid units, and by establishing simulation models, calculating the optimal path and cost, and combining coverage analysis, we can comprehensively evaluate the path quality and coverage quality of grid units, ensuring that the area division can meet the accessibility requirements of efficient operation.

[0109] S2. Obtain historical inbound and outbound information from the warehouse, use a long short-term memory network to adjust the boundaries of the inbound buffer zone, dynamic picking zone, and outbound temporary storage zone, and use an inbound and outbound coordination and consistency control algorithm to allocate tasks and coordinate the path scheduling of the cargo handling vehicles.

[0110] As a preferred implementation, the steps of acquiring historical warehouse inbound and outbound information, adjusting the boundaries of the inbound buffer zone, dynamic picking zone, and outbound temporary storage zone using a long short-term memory network, and allocating tasks and coordinating path scheduling for cargo handling vehicles based on an inbound / outbound coordination and consistency control algorithm include the following steps:

[0111] S21. Using a long short-term memory network, construct an inbound / outbound prediction model, and train the inbound / outbound prediction model using historical warehouse inbound and outbound information.

[0112] It should be noted that Long Short-Term Memory (LSTM) is a type of Recurrent Neural Network (RNN). By introducing gating mechanisms (input gate, forget gate, and output gate), it solves the gradient vanishing and gradient explosion problems that traditional RNNs encounter when processing long sequence data, and is able to better capture long-term dependencies in sequence data.

[0113] When building the model, the structure of the input layer, hidden layer, and output layer is determined and configured. The input layer receives historical warehouse inbound and outbound information, which can be daily, hourly, or shorter time interval data arranged in chronological order. The hidden layer contains multiple LSTM units to extract features and patterns from the data. The output layer outputs the predicted inbound and outbound quantities.

[0114] Then, the model is trained using historical warehouse inbound and outbound information. During training, the historical data is divided into a training set and a validation set. The training set is used to adjust the model's parameters so that it can better fit the historical data; the validation set is used to evaluate the model's performance and prevent overfitting. The model's parameters are continuously optimized iteratively until the model's performance on the validation set reaches a satisfactory level.

[0115] S22. Based on the trained inbound and outbound prediction model, predict the inbound and outbound quantities of the warehouse, and calculate the demand deviation based on the predicted inbound and outbound quantities.

[0116] It should be noted that demand deviation refers to the difference between the predicted inbound and outbound quantities and the actual demand. For example, if the goal of a warehouse is to maintain a certain inventory level, the required inventory level for a future period can be calculated based on the current inventory and the predicted inbound and outbound quantities. This level can then be compared with the actual inventory to obtain the demand deviation.

[0117] S23. Based on demand deviation and combined with preset deviation thresholds, adjust the distinction boundaries of the inbound buffer zone, dynamic picking zone and outbound temporary storage zone respectively.

[0118] Specifically, deviation thresholds are set based on warehouse operational strategies and actual conditions to determine whether demand deviations necessitate adjustments to area boundaries. For example, a pre-set deviation threshold for the inbound buffer zone might be 10% of the predicted inbound volume versus actual demand. If the actual inbound volume exceeds the forecast by more than 10%, the area is considered insufficient to handle the upcoming inbound volume, and the buffer zone's boundaries are adjusted to expand its area and accommodate more goods. Similarly, when there is a significant deviation between the predicted outbound volume and actual demand, the boundaries of the outbound temporary storage area also need adjustment. Through this dynamic area adjustment, the warehouse can flexibly optimize space allocation based on actual demand changes, improve resource utilization, reduce bottlenecks and congestion, and ensure efficient and flexible warehouse management. If the deviation is not greater than the threshold, there is no need to adjust the boundaries distinguishing the inbound buffer zone, dynamic picking area, and outbound temporary storage area.

[0119] S24. Based on the predicted inbound and outbound volumes, the inbound and outbound coordination consistency control algorithm is used to allocate tasks and coordinate the path scheduling of the cargo handling vehicles.

[0120] As a preferred embodiment, the step of allocating tasks and coordinating path scheduling for cargo handling vehicles based on predicted inbound and outbound volumes using an inbound / outbound coordination consistency control algorithm includes the following steps:

[0121] S241. Based on the area information of the inbound buffer zone, dynamic picking zone, and outbound temporary storage zone, as well as the configuration information of the cargo handling vehicle, construct the warehouse-level intelligent agent and the cargo handling vehicle-level intelligent agent.

[0122] Specifically, the inbound buffer zone information includes its location, capacity, and storage method; the dynamic picking zone information includes shelf layout, product type distribution, and picking aisle conditions; and the outbound temporary storage zone information includes its location, capacity, and stacking rules. The configuration information for the freight forwarding vehicles includes their quantity, model, load capacity, speed, and range.

[0123] The warehouse-level intelligent agent is a macro-level decision-making entity responsible for coordinating the overall operation of the warehouse. It can acquire information from various areas and the configuration information of transport vehicles, coordinating and allocating tasks from a global perspective to ensure the overall operational efficiency of the warehouse.

[0124] The transport vehicle-level intelligent agent is a micro-level decision-making entity. Each transport vehicle is equipped with a transport vehicle-level intelligent agent. Based on the tasks issued by the warehouse-level intelligent agent, the transport vehicle-level intelligent agent combines its own status (such as current location, remaining battery power, load status, etc.) and the real-time environmental information of the warehouse (such as aisle congestion, obstacle location, etc.) to perform specific task execution and path planning.

[0125] S242. Based on the predicted inbound and outbound volumes, determine the target task volume for the inbound buffer zone, dynamic picking zone, and outbound temporary storage zone, respectively.

[0126] Specifically, if it's predicted that 1000 items will be received in the next day, and the inbound buffer can handle 100 items per hour, then the target workload for the inbound buffer can be set to process a certain number of items per hour. If it's predicted that there will be 500 orders in the next day, and each order requires picking an average of 5 items, then the target workload for the dynamic picking area is 2500 items picked per day. If it's predicted that 800 items will be shipped out in the next day, and the outbound staging area can handle 80 items per hour, then the target workload for the outbound staging area can be set to process a certain number of items per hour to ensure that the goods can be shipped out smoothly.

[0127] S243. Based on the target task volume of the inbound buffer, dynamic picking area and outbound temporary storage area, the warehouse-level intelligent agent coordinates the task requirements among the inbound buffer, dynamic picking area and outbound temporary storage area through a consistency control algorithm.

[0128] As a preferred embodiment, the warehouse-level agent coordinates the task requirements among the inbound buffer, dynamic picking area, and outbound temporary storage area based on the target task volume of the inbound buffer, dynamic picking area, and outbound temporary storage area through a consistency control algorithm, including the following steps:

[0129] S2431. The target task volume of the inbound buffer, dynamic picking area and outbound temporary storage area is modeled as an equivalent task load.

[0130] Specifically, equivalent task load comprehensively considers factors such as the number, complexity, and processing time of tasks. For example, for the inbound buffer, the inbound operation of each item can be regarded as a basic task unit. The load coefficient of each task unit is determined based on factors such as the type of goods and packaging method. Then, the target inbound quantity is multiplied by the load coefficient to obtain the equivalent task load of the inbound buffer. Similarly, this method is used for modeling dynamic picking areas and outbound temporary storage areas.

[0131] S2432. Construct a communication topology diagram between the inbound buffer, dynamic picking area and outbound temporary storage area, and construct an adjacency matrix based on the inter-region task migration and resource sharing relationship;

[0132] It should be noted that the inbound buffer zone, dynamic picking zone, and outbound temporary storage zone need to exchange information and coordinate tasks, therefore a communication topology diagram needs to be constructed between them. The communication topology diagram describes the connection relationships between each area. For example, there is a goods handling connection between the inbound buffer zone and the dynamic picking zone, and there is also a goods handling connection between the dynamic picking zone and the outbound temporary storage zone.

[0133] The adjacency matrix is ​​a square matrix whose elements represent the connection strength between regions. If there is a direct task migration or resource sharing relationship between two regions, the element at the corresponding position in the adjacency matrix has a positive value, representing the connection strength; otherwise, the element has a value of 0.

[0134] S2433. Using the discrete consistency iterative formula, the task demand growth rate of the warehouse-level intelligent agent is coordinated to achieve consistency through the warehouse-level intelligent agent.

[0135] It should be noted that the discrete consistency iterative formula is as follows:

[0136] ;

[0137] In the formula, C ( t ) is the task increment state vector for all regions; α This is the step size adjustment coefficient; T To control latency, this process can be run iteratively until each... Ci ( t Convergence and consistency mean that the marginal task execution cost is consistent across all regions. Once the incremental rate converges, the task quantity of each region can be solved in reverse according to its incremental rate. These task quantities serve as the final "regional task allocation instructions" and are issued to the transport vehicle-level agents in each region for subsequent path and task execution scheduling.

[0138] S244. The warehouse-level agent distributes the coordinated task requirements to the transport vehicle-level agents in the inbound buffer, dynamic picking area, and outbound temporary storage area, and optimizes the task allocation and path planning of each transport vehicle through a consistency control algorithm.

[0139] It should be noted that when multiple transport vehicles are performing tasks in the same area, the transport vehicle-level intelligent agent can coordinate its own task allocation based on the task status and location information of other transport vehicles through a consistency control algorithm to avoid task conflicts and duplicate operations. For example, if two transport vehicles have received the task of going to the same shelf to pick up goods, the transport vehicle-level intelligent agent can reallocate the task through the consistency control algorithm, so that one transport vehicle goes to the shelf and the other transport vehicle performs other tasks.

[0140] Meanwhile, the vehicle-level intelligent agent also utilizes a consensus control algorithm to optimize path planning. In a warehouse environment, there may be multiple paths to the task location. The vehicle-level intelligent agent can select the optimal path based on the travel paths of other vehicles and real-time traffic conditions through a consensus control algorithm. For example, if a passageway becomes congested, the vehicle-level intelligent agent can adjust its own path in a timely manner by interacting with other vehicles to avoid congested areas and improve handling efficiency. In this way, it can be ensured that each goods handling vehicle can complete its task efficiently, improving the overall operational efficiency of the warehouse.

[0141] Specifically, in terms of task allocation and path coordination scheduling for cargo handling vehicles, warehouse-level and vehicle-level intelligent agents are constructed. Based on the predicted task volume, the target task volume for each area is determined. The warehouse-level intelligent agent coordinates the task requirements between areas with the help of a consistency control algorithm, and distributes the coordinated tasks to the vehicle-level intelligent agent to optimize task allocation and path planning. This improves the operating efficiency of cargo handling vehicles, reduces operating costs, enhances the coordination and flexibility of the overall warehouse operation, and helps to improve the intelligence level and overall efficiency of warehouse management.

[0142] S3. Utilize digital twin technology to construct a five-dimensional twin model of the warehouse, and use the five-dimensional twin model of the warehouse to continuously monitor, adjust and optimize the inbound and outbound operation processes in order to achieve collaborative scheduling of inbound and outbound operations.

[0143] It should be noted that digital twin technology establishes a digital mapping of physical entities in virtual space, and uses real-time data to drive the virtual model to run synchronously with the physical entity, thereby enabling status monitoring, performance analysis and optimization decisions for the physical entity. Compared with the traditional three-dimensional model, the five-dimensional twin model of the warehouse adds the time dimension and the functional dimension, which can more comprehensively and dynamically reflect the actual situation of the warehouse.

[0144] First, when constructing the five-dimensional twin model of the warehouse, 3D modeling software (such as AutoCAD, 3ds Max, etc.) is used to accurately model the warehouse's architectural structure, shelving layout, aisle settings, etc., including the warehouse's length, width, and height dimensions, the number of shelving layers and spacing, and the width and direction of aisles, to intuitively display the warehouse's spatial form. Then, various equipment (such as freight vehicles, stacker cranes, conveyor belts, etc.) and goods within the warehouse are digitally modeled, recording the equipment's model, specifications, performance parameters, and the goods' type, size, weight, and other information, enabling the virtual model to accurately reflect the physical objects within the warehouse.

[0145] Secondly, real-time data from the warehouse (such as goods entry and exit times, equipment operating times, and personnel operation times) is linked to the virtual model, enabling the model to update its status in real time as time changes. Various sensors (such as RFID readers, photoelectric sensors, and temperature and humidity sensors) are installed in the warehouse to collect real-time data on the location, quantity, and status of goods; the operating status of equipment (such as speed, load, and fault information); and environmental parameters (such as temperature, humidity, and light intensity).

[0146] Specifically, a journal in the field of Computer Integrated Manufacturing Systems published an article titled "Digital Twin Five-Dimensional Model and Applications in Ten Major Fields." This article describes the design of automated warehouses based on digital twins. It involves establishing five-dimensional digital twin models of various equipment within the automated warehouse, utilizing a design demonstration platform to achieve near-physical semi-physical simulation design. The platform allows for 3D image design of the warehouse layout and semi-physical simulation verification based on equipment such as shelving, transportation equipment, and robots. It also completes geometric modeling, action script writing, and the definition of instruction and information interfaces, achieving modular encapsulation and customized model interface design. Leveraging the five-dimensional model of the automated warehouse and its equipment, a remote operation and maintenance service platform is built for users. This enables remote operation and maintenance of the automated warehouse based on digital twins. By establishing a virtual model that fully maps to the automated warehouse, and utilizing the warehouse's data and various algorithms, real-time simulation and optimization of the automated warehouse can be achieved. While monitoring the warehouse's real-time status and information, functions such as inventory management, location management, cost management, early warning management, predictive maintenance, and job scheduling are provided as software services to users with different needs.

[0147] Therefore, in this invention, the five-dimensional twin model is integrated with other warehouse management systems (such as Warehouse Management System (WMS) and Transportation Management System (TMS)) to achieve information sharing and collaboration. Based on the warehouse's real-time operational status and inbound / outbound operation requirements, the five-dimensional twin model can dynamically allocate various resources within the warehouse, such as equipment, personnel, and storage space.

[0148] Specifically, by utilizing a five-dimensional twin model of the warehouse, operational data can be collected in real time through sensors and equipment on the physical entity side. A three-dimensional visualization scene can be constructed on the virtual model side, and the process can be dynamically simulated. By combining the twin data dimensions to analyze multi-source data to explore optimization potential, intelligent scheduling algorithms and automated adjustment strategies can be developed based on the service dimension. Finally, the connection dimension ensures real-time data flow and system collaboration, forming a closed loop of "data-driven - simulation prediction - decision optimization - feedback iteration". This enables dynamic monitoring, path optimization, equipment scheduling adjustment, and long-term efficiency improvement of inbound and outbound operations. For example, when there are many inbound and outbound tasks in a certain area, the model can automatically allocate more handling vehicles and personnel to that area to improve operational efficiency.

[0149] like Figure 2 As shown, according to an embodiment of the present invention, an intelligent warehouse management system that can improve the efficiency of warehouse management and scheduling is provided. The system includes: a region division module 1, a collaborative scheduling module 2, and a job adjustment module 3, and the region division module 1, the collaborative scheduling module 2, and the job adjustment module 3 are connected in sequence.

[0150] The area division module 1 is used to obtain warehouse storage information. Based on the spatial gridding method and combined with the accessibility analysis method, the warehouse is divided into an inbound buffer zone, a dynamic picking zone, and an outbound temporary storage zone.

[0151] The collaborative scheduling module 2 is used to obtain historical warehouse inbound and outbound information, use a long short-term memory network to adjust the regional boundaries of the inbound buffer zone, dynamic picking zone and outbound temporary storage zone, and perform task allocation and path collaborative scheduling for cargo handling vehicles based on the inbound and outbound coordination consistency control algorithm.

[0152] The operation adjustment module 3 is used to build a five-dimensional twin model of the warehouse using digital twin technology. The five-dimensional twin model of the warehouse is used to continuously monitor, adjust and optimize the inbound and outbound operation processes in order to achieve collaborative scheduling of inbound and outbound operations.

[0153] Those skilled in the art will understand that embodiments of the present invention can be provided as methods, systems, or computer program products. Therefore, the present invention can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, the present invention can take the form of a computer program product embodied on one or more computer-usable storage media (including, but not limited to, disk storage, optical storage, etc.) containing computer-usable program code.

[0154] The specific embodiments described above further illustrate the purpose, technical solution, and beneficial effects of the present invention. It should be understood that the above descriptions are merely specific embodiments of the present invention and are not intended to limit the scope of protection of the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the scope of protection of the present invention.

Claims

1. An intelligent warehouse management method that can improve the efficiency of warehouse management and scheduling, characterized in that, Includes the following steps: S1. Obtain the storage space area of ​​the warehouse, perform grid processing on the storage space area, and determine the location information of each grid cell based on the coordinate transformation method; The storage capacity, cargo turnover characteristics, and related operational requirements of each grid cell are analyzed, and functional attribute labels are configured for each grid cell. Based on the location information of each grid cell, accessibility analysis is used to simulate the operational process within the warehouse, and accessibility analysis is performed on each grid cell to obtain the accessibility analysis results. Based on the functional attribute labels and accessibility analysis results, the grid cells are clustered into inbound buffer zone, dynamic picking zone, and outbound temporary storage zone; S2. Obtain historical inbound and outbound information from the warehouse, use a long short-term memory network to adjust the boundaries of the inbound buffer zone, dynamic picking zone, and outbound temporary storage zone, and use an inbound and outbound coordination and consistency control algorithm to allocate tasks and coordinate the path scheduling of the cargo handling vehicles. The process of simulating the warehouse workflow based on the location information of each grid cell using accessibility analysis, and performing accessibility analysis on each grid cell to obtain the accessibility analysis results includes the following steps: Based on the structural characteristics of the warehouse, a simulation model of the warehouse is established using the finite element analysis method, and the warehouse entrance is taken as the starting point of the path. Starting from the path origin, the operation process within the warehouse is simulated, and a path search algorithm is used to calculate the optimal path from the path origin to each grid cell and its corresponding path cost, thus obtaining the path calculation results. Based on the path calculation results, and using the coverage analysis method, accessibility is evaluated for each grid cell to obtain accessibility analysis results; The process of evaluating accessibility for each grid cell based on path calculation results and using coverage analysis to obtain accessibility analysis results includes the following steps: Centered on the inlet, the evaluation of each grid cell is carried outward layer by layer, and the path calculation results of each grid cell in each layer are collected. Based on the distribution of obstacles in the warehouse and the information of the cargo handling vehicles, the occlusion angle corresponding to each grid cell is calculated, and the path quality of each grid cell is evaluated in combination with the path cost. Based on the different work processes of each grid cell, the coverage of each grid cell is analyzed to determine the coverage quality of each grid cell; Based on the path quality and coverage quality of each grid cell, accessibility is assessed for each grid cell, and accessibility analysis results are generated.

2. The intelligent warehouse management method according to claim 1, which can improve the efficiency of warehouse management and scheduling, is characterized in that, The analysis of the storage capacity, cargo turnover characteristics, and related operational requirements of each grid cell, and the configuration of functional attribute tags for each grid cell, includes the following steps: Collect the structural information and static attribute data of each grid cell, and combine them with the historical inbound and outbound records of each grid cell to analyze the historical cargo storage type, storage frequency and turnover cycle of each grid cell; Based on structural information, static attribute data, and historical cargo storage conditions, the fuzzy comprehensive evaluation method is used to assess the storage capacity of each grid unit, and cargo turnover characteristics are summarized based on cargo turnover cycle and storage frequency. Based on the assessment results, the storage capacity, cargo turnover characteristics and operational requirements of each grid unit are correlated and matched to determine the functional type of each grid unit; Configure a functional attribute label for each grid cell based on its functional type.

3. The intelligent warehouse management method according to claim 2, which can improve the efficiency of warehouse management and scheduling, is characterized in that, The formula for calculating the shading angle corresponding to each grid cell based on the obstacle distribution and cargo handling vehicle information within the warehouse is as follows: ； In the formula, θ obs Indicates the shading angle. h obs Indicates the height of the obstacle's vertex. h AGV This indicates the height of the top of the cargo handling vehicle. d ij Represents grid cells i To the observation point j Horizontal distance λ Indicates the dynamic obstacle sensitivity coefficient. δ dynamic This indicates the amount of dynamic obstacle correction.

4. The intelligent warehouse management method according to claim 1, which can improve the efficiency of warehouse management and scheduling, is characterized in that, The process of combining functional attribute tags and accessibility analysis results to cluster grid cells into inbound buffer zones, dynamic picking zones, and outbound temporary storage zones includes the following steps: Based on the functional attribute labels and accessibility analysis results, a clustering index system is constructed; The inbound buffer zone, dynamic picking zone, and outbound temporary storage zone are selected as target categories, and the number of clusters is configured accordingly. The number of cluster centers is selected randomly from all grid cells. Calculate the distance from each grid cell to the initial cluster center, and assign the grid cells to the corresponding target categories according to the nearest distance principle; Recalculate the cluster center for each target category, and repeat the process of calculating distance and assigning categories until the cluster centers no longer change, thus obtaining the final clustering result.

5. The intelligent warehouse management method according to claim 1, which can improve the efficiency of warehouse management and scheduling, is characterized in that, The process of acquiring historical warehouse inbound and outbound information, adjusting the boundaries of the inbound buffer zone, dynamic picking zone, and outbound temporary storage zone using a long short-term memory network, and allocating tasks and coordinating path scheduling for cargo handling vehicles based on an inbound / outbound coordination and consistency control algorithm includes the following steps: S21. Using a long short-term memory network, construct an inbound / outbound prediction model, and train the inbound / outbound prediction model using historical warehouse inbound and outbound information. S22. Based on the trained inbound and outbound prediction model, predict the inbound and outbound quantities of the warehouse, and calculate the demand deviation based on the predicted inbound and outbound quantities. S23. Based on demand deviation and combined with preset deviation thresholds, adjust the distinction boundaries of the inbound buffer zone, dynamic picking zone and outbound temporary storage zone respectively. S24. Based on the predicted inbound and outbound volumes, the inbound and outbound coordination and consistency control algorithm is used to allocate tasks and coordinate the path scheduling of the cargo handling vehicles.

6. The intelligent warehouse management method according to claim 5, which can improve the efficiency of warehouse management and scheduling, is characterized in that, The process of allocating tasks and coordinating paths for cargo handling vehicles based on predicted inbound and outbound volumes, using an inbound / outbound coordination and consistency control algorithm, includes the following steps: S241. Based on the area information of the inbound buffer zone, dynamic picking zone, and outbound temporary storage zone, as well as the configuration information of the cargo handling vehicle, construct the warehouse-level intelligent agent and the cargo handling vehicle-level intelligent agent. S242. Based on the predicted inbound and outbound volumes, determine the target task volume for the inbound buffer zone, dynamic picking zone, and outbound temporary storage zone, respectively. S243. Based on the target task volume of the inbound buffer, dynamic picking area and outbound temporary storage area, the warehouse-level intelligent agent coordinates the task requirements among the inbound buffer, dynamic picking area and outbound temporary storage area through a consistency control algorithm. S244. The warehouse-level agent distributes the coordinated task requirements to the transport vehicle-level agents in the inbound buffer, dynamic picking area and outbound temporary storage area, and optimizes the task allocation and path planning of each transport vehicle through the consistency control algorithm. Based on the target task volume of the inbound buffer, dynamic picking area, and outbound temporary storage area, the warehouse-level intelligent agent coordinates the task requirements among the inbound buffer, dynamic picking area, and outbound temporary storage area through a consistency control algorithm, including the following steps: S2431. The target task volume of the inbound buffer, dynamic picking area and outbound temporary storage area is modeled as an equivalent task load. S2432. Construct a communication topology diagram between the inbound buffer, dynamic picking area and outbound temporary storage area, and construct an adjacency matrix based on the inter-region task migration and resource sharing relationship; S2433. Using the discrete consistency iterative formula, the task demand growth rate of the warehouse-level intelligent agent is coordinated to achieve consistency through the warehouse-level intelligent agent.

7. The intelligent warehouse management method according to claim 6, which can improve the efficiency of warehouse management and scheduling, is characterized in that, The process of acquiring historical warehouse inbound and outbound information, adjusting the boundaries of the inbound buffer zone, dynamic picking zone, and outbound temporary storage zone using a long short-term memory network, and allocating tasks and coordinating path scheduling for cargo handling vehicles based on an inbound / outbound coordination and consistency control algorithm includes: A five-dimensional twin model of the warehouse is constructed using digital twin technology. This model is then used to continuously monitor, adjust, and optimize the inbound and outbound operations within the warehouse, thereby achieving coordinated scheduling of inbound and outbound processes.

8. An intelligent warehouse management system for improving warehouse management and scheduling efficiency, used to implement the intelligent warehouse management method for improving warehouse management and scheduling efficiency as described in any one of claims 1-7, characterized in that, The system includes: The area division module is used to acquire the storage space areas of the warehouse, grid the storage space areas, and determine the location information of each grid cell based on the coordinate transformation method; analyze the storage capacity, cargo turnover characteristics, and related operational requirements of each grid cell, and configure functional attribute labels for each grid cell; based on the location information of each grid cell, use accessibility analysis to simulate the operation process within the warehouse, and perform accessibility analysis on each grid cell to obtain accessibility analysis results; combining the functional attribute labels and accessibility analysis results, the grid cells are clustered into inbound buffer zones, dynamic picking zones, and outbound temporary storage zones; The collaborative scheduling module is used to obtain historical warehouse inbound and outbound information, use a long short-term memory network to adjust the regional boundaries of the inbound buffer zone, dynamic picking zone and outbound temporary storage zone, and perform task allocation and path collaborative scheduling for cargo handling vehicles based on the inbound and outbound coordination and consistency control algorithm.

Images