VMI warehouse inventory dynamic regulation method and system based on industrial internet

Abstract

The application provides an industrial internet-based VMI warehouse inventory dynamic regulation method and system, which comprises the following steps: constructing a warehouse digital twin model based on the warehouse state information of a VMI entity warehouse; constructing an initial DOM task structure tree according to a plurality of sub-tasks obtained by disassembling received material warehouse-in and warehouse-out order information according to the logical sequence of task execution; simulating the execution process of the material warehouse-in and warehouse-out task in the warehouse digital twin model based on the initial DOM task structure tree, obtaining a conflict task with an abnormality; performing task optimization based on the task node and task type of the conflict task, obtaining an optimized execution strategy, updating the initial execution strategy corresponding to the conflict task in the initial DOM task structure tree based on the optimized execution strategy, and obtaining a target DOM task structure tree; and performing warehouse inventory regulation on the material warehouse-in and warehouse-out in the VMI entity warehouse based on the target DOM task structure tree. The application improves the overall warehouse operation efficiency.

Description

Technical Field

[0001] This invention relates to the field of industrial service technology, and in particular to a method and system for dynamic control of VMI warehouse inventory based on the Industrial Internet. Background Technology

[0002] In today's warehouse management field, efficient control of material inbound and outbound operations is crucial. As warehouse scale expands and business complexity increases, traditional inventory control methods are increasingly inadequate. Some existing methods arrange material inbound and outbound operations based on a simple order sequence. This approach fails to adequately consider the real-time status of the warehouse and the relationships between materials, leading to frequent material backlogs or stockouts in practice. For example, when multiple orders are placed simultaneously with similar demand times for different materials, arranging inbound according to order sequence may result in cumbersome inbound processes for some materials, consuming excessive warehouse resources and preventing other urgently needed materials from being inbound in a timely manner. This, in turn, affects subsequent outbound operations and reduces overall warehouse operational efficiency. Summary of the Invention

[0003] This invention provides a method and system for dynamic control of warehouse inventory based on the Industrial Internet, aiming to solve the problem of frequent material backlog or shortage in actual operation and improve the overall efficiency of warehouse operation.

[0004] In a first aspect, the present invention provides a VMI (Vendor Managed Inventory) dynamic control method for warehouses based on the Industrial Internet, comprising:

[0005] Based on the collected storage status information of the VMI physical warehouse, a storage digital twin model is constructed that is proportionally mapped to the VMI physical warehouse.

[0006] The received material inbound / outbound order information is broken down into multiple sub-tasks according to the order of task execution, and an initial DOM task structure tree is constructed based on the multiple sub-tasks.

[0007] Based on the initial DOM task structure tree, the execution process of material inbound and outbound tasks is simulated in the warehouse digital twin model to obtain conflicting tasks that are abnormal during the simulation.

[0008] Based on the task nodes and task types of the conflicting tasks, task optimization is performed to obtain an optimized execution strategy. Based on the optimized execution strategy, the initial execution strategy corresponding to the conflicting task is updated in the initial DOM task structure tree to obtain the target DOM task structure tree.

[0009] Based on the target DOM task structure tree, the warehouse inventory is adjusted for the inbound and outbound of materials in the VMI entity warehouse.

[0010] Secondly, the present invention also provides a VMI warehouse inventory dynamic control system based on the Industrial Internet, including: a twin model construction module, a structure tree construction module, a task simulation module, a strategy update module, and a warehouse inventory control module;

[0011] The twin model construction module is used to construct a warehouse digital twin model that is proportionally mapped to the VMI physical warehouse based on the collected warehouse status information.

[0012] The structure tree construction module is used to break down the received material inbound and outbound order information according to the order of task execution, obtain multiple sub-tasks, and construct an initial DOM task structure tree based on the multiple sub-tasks.

[0013] The task simulation module is used to simulate the execution process of material entry and exit tasks in the warehouse digital twin model based on the initial DOM task structure tree, and to obtain conflicting tasks that are abnormal during the simulation process.

[0014] The strategy update module is used to optimize the task based on the task node and task type of the conflicting task to obtain the optimized execution strategy, and update the initial execution strategy corresponding to the conflicting task in the initial DOM task structure tree based on the optimized execution strategy to obtain the target DOM task structure tree.

[0015] The warehouse inventory control module is used to control the inbound and outbound of materials in the VMI entity warehouse based on the target DOM task structure tree.

[0016] Thirdly, the present invention also provides an electronic device, comprising: a memory for storing computer software programs; and a processor for reading and executing the computer software programs, thereby realizing the VMI warehouse inventory dynamic control method based on the Industrial Internet as described above.

[0017] Fourthly, the present invention also provides a non-transitory computer-readable storage medium storing a computer software program, which, when executed by a processor, implements the VMI warehouse inventory dynamic control method based on the Industrial Internet as described above.

[0018] Fifthly, the present invention also provides a computer program product, including a computer program, which, when executed by a processor, implements the above-mentioned VMI warehouse inventory dynamic control method based on the Industrial Internet.

[0019] The VMI (Vendor Managed Inventory) dynamic control method based on the Industrial Internet provided in this invention constructs a warehouse digital twin model that is proportionally mapped to the VMI physical warehouse. This model reflects the real-time status of the VMI physical warehouse, avoiding decision-making errors caused by untimely understanding of the warehouse status. An initial DOM (Domain-Oriented Task) structure tree is constructed based on multiple sub-tasks in the material inbound / outbound order information, clearly presenting the task execution logic and facilitating the detection of conflicting tasks. The initial DOM structure tree is then used for pre-simulation in the warehouse digital twin model. Task optimization is performed based on the task nodes and task types of conflicting tasks that are abnormal during the simulation, resulting in a target DOM structure tree. Finally, warehouse inventory control is performed on the material inbound / outbound operations in the VMI physical warehouse based on the target DOM structure tree. Therefore, through pre-simulation and optimization, conflicts in physical operations are avoided, ensuring smooth task execution during physical operations. This solves the problem of frequent material backlogs or stockouts in actual operations, reduces equipment waiting time and inefficient handling, improves material inbound / outbound efficiency, and thus improves overall warehouse operation efficiency. Attached Figure Description

[0020] Figure 1 This is a flowchart of the VMI warehouse inventory dynamic control method based on the Industrial Internet provided in this embodiment of the invention;

[0021] Figure 2 This is a structural diagram of the VMI warehouse inventory dynamic control system based on the Industrial Internet provided in this embodiment of the invention;

[0022] Figure 3 An embodiment diagram of the electronic device provided in this invention;

[0023] Figure 4 An embodiment diagram of a computer-readable storage medium provided in accordance with the present invention. Detailed Implementation

[0024] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0025] Optional, see below Figure 1 , Figure 1 This is a flowchart of the VMI (Vendor Managed Inventory) dynamic control method for warehouse inventory based on the Industrial Internet provided by this invention. In this embodiment of the invention, the executing entity of the VMI dynamic control method for warehouse inventory based on the Industrial Internet is the warehouse inventory control system. Therefore, the VMI dynamic control method for warehouse inventory based on the Industrial Internet includes:

[0026] Step 10: Based on the collected storage status information of the VMI physical warehouse, construct a warehouse digital twin model that is proportionally mapped to the VMI physical warehouse.

[0027] Optionally, the warehouse inventory control system collects warehouse status information from VMI (Vendor Managed Inventory) physical warehouses. This information includes the warehouse's spatial layout (such as the location, quantity, and size of shelves, aisle width and direction), equipment information (such as forklift models, quantity, location, and operating status, conveyor belt length, location, and operating parameters), material information (such as material type, quantity, storage location, and packaging specifications), and environmental information (such as temperature, humidity, and lighting). Furthermore, based on the collected warehouse status information, the system constructs a corresponding digital model in virtual space according to a proportional scaling principle. This ensures that each element in the model corresponds one-to-one with the elements in the physical warehouse, and that their location, size, and attributes are completely consistent. A real-time data interaction mechanism is established between the physical warehouse and the digital twin model, enabling the digital twin model to synchronize with the physical warehouse's status changes in real time.

[0028] In one embodiment, a VMI physical warehouse is a warehouse that is 100 meters long, 50 meters wide, and 10 meters high, with 50 shelves inside. Each shelf is 5 meters long, 1 meter wide, and 3 meters high, and is divided into 5 layers. There are 10 forklifts, all of model XYZ-100, of which 8 are in normal operation and 2 are charging. The stored materials are of three types: A, B, and C. Material A consists of 100 boxes, each 0.5 meters long, 0.5 meters wide, and 0.5 meters high, stored on the first layer of shelves 1-10; Material B consists of 200 boxes, stored on the second layer of shelves 11-30; and Material C consists of 150 boxes, stored on the third layer of shelves 31-50. The aisles inside the warehouse are 3 meters wide and run east-west. The ambient temperature is maintained at 25°C, and the humidity is maintained at 50%.

[0029] Furthermore, based on all the above information, the warehouse inventory control system constructs a virtual digital twin model of the warehouse in a 1:1 scale. In the model, the location and size of the shelves are completely consistent with the physical warehouse, the location and operating status of the forklifts are synchronized in real time, and the storage location and quantity of materials also correspond to the physical warehouse. For example, if there are 10 boxes of material A on the first level of shelf 5 in the physical warehouse, the digital twin model will also show 10 boxes of material A on the first level of shelf 5. When a forklift in the physical warehouse moves from aisle 3 to near shelf 5, the corresponding forklift in the digital twin model will also move to the same position synchronously.

[0030] Step 20: Decompose the received material inbound / outbound order information according to the order of task execution to obtain multiple sub-tasks, and construct an initial DOM task structure tree based on the multiple sub-tasks.

[0031] Furthermore, after receiving material inbound / outbound order information, the warehouse inventory control system breaks it down into multiple sub-tasks according to the order of task execution. Each sub-task has clear attributes, including the type of pre-constraints (i.e., which other sub-tasks must be completed before the sub-task can be executed), resource requirements (such as the number of forklifts, operators, specific shelf locations, etc. required), and execution time window (i.e., the time range within which the sub-task can start execution and must be completed).

[0032] Meanwhile, there are multiple task relationships between subtasks, including sequential dependency, parallel compatibility, and branch triggering. Sequential dependency reflects the mandatory order of task execution; for example, material receiving and shelving must be completed before material picking can proceed. Parallel compatibility reflects the resource and time compatibility of tasks that can be executed simultaneously; for example, two receiving subtasks running in different areas can be executed in parallel if their required resources do not conflict and their time windows overlap. Branch triggering reflects the fulfillment of conditions for a task to serve as a branch starting point; for example, when the quantity of received materials exceeds a certain threshold, a branch to add a shelf-organizing subtask is triggered.

[0033] Furthermore, the warehouse inventory control system constructs an initial DOM task structure tree based on the attributes of each subtask and the task relationships between subtasks, as detailed in steps 201 to 205.

[0034] In one embodiment, the warehouse inventory control system receives a material inbound / outbound order: to inbound 20 boxes of material A and simultaneously outbound 10 boxes of material B. The order is broken down as follows: Subtask 1: Unload 20 boxes of material A from the transport vehicle. Prerequisite constraint type: none (no other tasks required beforehand). Resource requirements: 1 forklift, 2 operators. Execution time window: 9:00-9:30. Subtask 2: Perform quality inspection on the 20 boxes of material A. Prerequisite constraint type: completion of Subtask 1. Resource requirements: 1 quality inspector. Execution time window: 9:20-9:50. Subtask 3: Move the 20 boxes of quality-inspected material A to shelves 1-10 and shelve them. Prerequisite constraint type: completion of Subtask 2. Resource requirements: 1 forklift, 1 operator. Execution time window: 9:40-10:10. Subtask 4: Pick 10 boxes of material B from shelves 11-30. Preconditions: None. Resource requirements: 1 forklift, 1 operator. Execution time window: 9:10-9:40. Subtask 5: Pack the 10 boxes of material B. Preconditions: Subtask 4 completed. Resource requirements: 1 packing operator. Execution time window: 9:30-10:00. Subtask 6: Load the 10 packed boxes of material B into the transport vehicle. Preconditions: Subtask 5 completed. Resource requirements: 1 forklift, 2 operators. Execution time window: 9:50-10:20.

[0035] Subtask Relationships: Sequential Dependency Dimension: Subtask 1 → Subtask 2 → Subtask 3; Subtask 4 → Subtask 5 → Subtask 6. Parallel Compatibility Dimension: Subtask 1 and Subtask 4 can be executed in parallel because the required forklifts operate in different areas, resources do not conflict, and the time windows 9:10-9:30 overlap. Branch Trigger Dimension: If more than 5 boxes of material A are found to be defective in subtask 2, then subtask 7 (moving the defective material A to the defective product area) is triggered. At this time, the pre-constraint type of subtask 7 is that the number of defective boxes in subtask 2 exceeds 5, the resource requirement is 1 forklift, and the execution time window is 9:50-10:20.

[0036] Based on the above subtasks and relationships, construct an initial DOM task structure tree. Each node in the tree represents a subtask, and the connections and labels between nodes reflect the task relationships and attributes.

[0037] Step 30: Based on the initial DOM task structure tree, simulate the execution process of material inbound and outbound tasks in the warehouse digital twin model to obtain conflicting tasks that are abnormal during the simulation.

[0038] Furthermore, the warehouse inventory control system maps the subtasks and their attributes and relationships in the initial DOM task structure tree to the warehouse digital twin model, simulating the execution process of each subtask, including resource allocation, time progression, and task coordination. During the simulation, the system monitors the task execution status in real time and identifies conflicting tasks with anomalies, as detailed in steps 301 to 303.

[0039] Optionally, the conflicting tasks in the embodiments of the present invention mainly include the following categories:

[0040] Time-sequence conflict task pair: refers to two subtasks that have a sequential dependency relationship, where the completion time of the first subtask exceeds the start time of the second subtask, resulting in a time-sequence conflict.

[0041] Conflicting parallel sibling task pairs: refers to sibling subtasks that should be executed in parallel, but cannot be executed simultaneously during simulation due to resource or time incompatibility.

[0042] Resource conflict task pair: refers to two subtasks competing for the same resource (such as a forklift or the same shelf location) during execution, resulting in the resource being unable to meet the needs of both tasks at the same time.

[0043] Continuing with the initial DOM task structure tree from the above embodiments, simulations are performed in the warehouse digital twin model:

[0044] Simulating the execution of subtask 1 (unloading material A from 9:00 to 9:30) and subtask 4 (picking material B from 9:10 to 9:40), it was found that both require the use of a forklift, but there is only one forklift available in the warehouse. At this time, subtask 1 and subtask 4 form a resource-occupancy conflict task pair.

[0045] The simulation includes subtask 2 (9:20-9:50 quality inspection of material A) and subtask 3 (9:40-10:10 shelving of material A). Subtask 3 is scheduled to start at 9:40, while subtask 2 may not be completed until 9:50. This results in subtask 3 starting earlier than subtask 2 completing, creating a time-coupling conflict between the two tasks.

[0046] Suppose there are also subtasks 6 (packing material B from 9:30 to 10:00) and 7 (a branch task that is supposed to be triggered, moving substandard material A from 9:50 to 10:20), both of which require one operator, but at this time only one operator is available, forming a resource-occupancy conflict task pair.

[0047] Step 40: Optimize the tasks based on the task nodes and task types of the conflicting tasks to obtain the optimized execution strategy. Then, update the initial execution strategy corresponding to the conflicting tasks in the initial DOM task structure tree based on the optimized execution strategy to obtain the target DOM task structure tree.

[0048] Furthermore, the warehouse inventory control system identifies the task nodes (i.e., which sub-tasks are conflicting) and task types of conflicting tasks. Task types include resource allocation conflict types (e.g., multiple tasks competing for the same resource), time planning conflict types (e.g., improper task scheduling), and branch coordination conflict types (e.g., coordination issues between branch tasks and the main task). The process of determining task types is detailed in steps A to D. Further, the warehouse inventory control system optimizes tasks based on the task nodes and task types to obtain an optimized execution strategy, as detailed in steps 401 to 417.

[0049] Optionally, for resource allocation conflicts, the conflict can be resolved by adding resources (e.g., temporarily calling a backup forklift) or adjusting the resource allocation scheme (e.g., staggering the use of the same resource at different times); for time planning conflicts, the execution time window of the task can be adjusted (e.g., postponing the start time of a task); for branch coordination conflicts, the branch triggering conditions can be redesigned or the execution plan of the branch task can be adjusted.

[0050] Furthermore, the warehouse inventory control system updates the initial execution strategy corresponding to the conflicting tasks in the initial DOM task structure tree based on the optimized execution strategy, such as modifying the execution time window, resource requirements, and pre-constraint types of subtasks, and finally obtains the target DOM task structure tree.

[0051] Continuing with the above embodiments, for the conflicting tasks identified in step 30:

[0052] For the resource conflict between subtask 1 and subtask 4 (resource allocation conflict type), the optimization strategy is to allocate one spare forklift from other warehouses. Subtask 1 will use the original forklift (9:00-9:30), and subtask 4 will use the spare forklift (9:10-9:40).

[0053] For the time conflict between subtask 2 and subtask 3 (time planning conflict type), adjust the execution time window of subtask 3 to 9:50-10:20 so that it starts after subtask 2 is completed.

[0054] For the resource conflict (resource allocation conflict type) between subtask 6 and subtask 7, adjust the execution time window of subtask 7 to 10:00-10:30. At this time, subtask 6 has been completed, and the operator can be free to execute subtask 7.

[0055] Based on these optimization strategies, the attributes and relationships of the corresponding subtasks in the initial DOM task structure tree are updated to obtain the target DOM task structure tree.

[0056] Step 50: Perform warehouse inventory control on material inbound and outbound operations in the VMI physical warehouse based on the target DOM task structure tree.

[0057] Furthermore, the warehouse inventory control system uses the target DOM task structure tree as a guide to control the execution of various material inbound and outbound operations in the physical warehouse. According to the sub-task sequence, resource allocation, and time windows specified in the target DOM task structure tree, it schedules relevant equipment and personnel to perform operations, such as directing forklifts to move materials according to planned routes and times, and arranging operators to perform quality inspection and packaging at designated times. During the control process, the execution status of each task is monitored in real time to ensure consistency with the requirements of the target DOM task structure tree. Simultaneously, based on the task execution results, the inventory information of the physical warehouse is updated promptly, such as increasing the inventory of incoming materials and decreasing the inventory of outbound materials.

[0058] Continuing with the above embodiments, based on the target DOM task structure tree obtained in step 40, the warehouse inventory control system controls the physical warehouse:

[0059] According to the optimized time and resource allocation, from 9:00 to 9:30, one existing forklift and two operators perform sub-task 1, unloading 20 boxes of material A; from 9:10 to 9:40, one spare forklift and one operator perform sub-task 4, picking 10 boxes of material B. From 9:20 to 9:50, one quality inspector performs sub-task 2, inspecting material A; from 9:50 to 10:20, one forklift and one operator perform sub-task 3, putting qualified material A on the shelves, at which point the inventory of material A increases by 20 boxes. 9:30-10:00, one packer performs sub-task 6, packing material B; 10:00-10:30, one operator performs sub-task 7 (if triggered), moving substandard material A; 10:00-10:30, one forklift and two operators perform the subsequent loading work of sub-task 6, loading 10 boxes of material B into the transport vehicle, at which point the inventory of material B decreases by 10 boxes.

[0060] This invention constructs a warehouse digital twin model that is proportionally mapped to the VMI physical warehouse, reflecting the real-time status of the VMI physical warehouse and avoiding decision-making errors caused by untimely understanding of the warehouse status. An initial DOM task structure tree is built based on multiple sub-tasks in the material inbound / outbound order information, clearly presenting the task execution logic and facilitating the detection of conflicting tasks. Then, the initial DOM task structure tree is used to perform pre-simulation in the warehouse digital twin model. Based on the task nodes and task types of conflicting tasks that appear abnormally during the simulation, task optimization is performed to obtain the target DOM task structure tree. Finally, warehouse inventory control is performed on the material inbound / outbound of the VMI physical warehouse based on the target DOM task structure tree. Therefore, through pre-simulation and optimization, conflicts in physical operations are avoided, ensuring the smooth execution of tasks in physical operations. This solves the problem of frequent material backlogs or stockouts in actual operations, reduces equipment waiting time and inefficient handling, improves material inbound / outbound efficiency, and thus improves overall warehouse operation efficiency.

[0061] In one embodiment, steps 201 to 205 include:

[0062] Step 201: Take the total task in the material inbound / outbound order information as the root node, and the subtask with no precondition type as the first subtask. Perform sequential dependency dimension matching between the first subtask and the root node to obtain the dependency dimension matching result.

[0063] Optionally, the warehouse inventory control system sets the overall task in the material inbound / outbound order information as the root node, which represents the overall goal of the entire inbound / outbound task. Further, the warehouse inventory control system selects subtasks from multiple subtasks that have no prerequisite constraints. These subtasks do not depend on any other subtasks and can serve as the starting point for the entire task execution, i.e., the first subtask.

[0064] Furthermore, the warehouse inventory control system performs sequential dependency dimension matching based on the first subtask and the root node. The sequential dependency dimension represents the mandatory relationship of tasks being executed in sequence. The matching in this embodiment mainly judges the rationality of the execution order between the root node and the first subtask, that is, whether the total task represented by the root node is taken as the starting point and the first subtask is taken as the first executable subtask under the total task, and whether it conforms to the sequential logic of task execution. After the matching is completed, the dependency dimension matching result is obtained.

[0065] In one embodiment, the overall task of the material inbound / outbound order is "to complete the inbound of 20 boxes of material A and the outbound of 10 boxes of material B", and this overall task is set as the root node. Among the decomposed subtasks, subtask 1 (unloading 20 boxes of material A from the transport vehicle) and subtask 4 (picking 10 boxes of material B from shelves 11-30) both have no prerequisites, so these two subtasks are determined as the first subtask.

[0066] Furthermore, the warehouse inventory control system performs sequential dependency matching between these two first subtasks and the root node. Since the root node represents the entire inbound and outbound task, and subtasks 1 and 4 are the initial operations under the overall task that do not depend on other subtasks, they conform to the logical order that the overall task is executed before the subtasks. Therefore, the dependency matching result is correct.

[0067] Step 202: If the dependency dimension matching result is consistent, determine whether there is a conflict between the resource requirements of the first subtask and the initial resource pool of the root node, and whether the execution time window of the first subtask includes the virtual start time of the root node. A consistent dependency dimension matching result indicates that the virtual execution time of the root node is earlier than the start of the execution time window of the first subtask, and the virtual execution completion status of the root node is in the completed state.

[0068] Furthermore, the warehouse inventory control system checks whether there is a conflict between the resource requirements of the first subtask and the initial resource pool of the root node. The initial resource pool of the root node refers to the initial set of resources allocated for the entire material inbound and outbound task, including forklifts, operators, quality inspection equipment, etc. If the resources required by the first subtask can be met in the initial resource pool and will not conflict with the allocation of other necessary initial resources, then there is no conflict in resource requirements; otherwise, there is a conflict. Further, the warehouse inventory control system checks whether the execution time window of the first subtask includes the virtual start time of the root node. The virtual start time of the root node is a manually set time point at which the entire overall task plan begins execution. The execution time window of the first subtask needs to include this virtual start time to ensure that the first subtask can start smoothly after the start time point of the overall task plan, conforming to the time logic of task execution. Continuing the above embodiment, the initial resource pool of the root node includes 2 forklifts, 3 operators, and 1 set of quality inspection equipment. The resource requirements of the first subtask 1 are 1 forklift and 2 operators, and the resource requirements of the first subtask 4 are 1 forklift and 1 operator. Upon inspection, the resources in the initial resource pool are sufficient to meet the needs of both subtasks, and there is no overlap or conflict between their resource requirements; therefore, there is no resource requirement conflict. The virtual start time of the root node is set to 9:00. The execution time window for subtask 1 is 9:00-9:30, inclusive of 9:00; the execution time window for subtask 4 is 9:10-9:40, also inclusive of 9:00 (the virtual start time is within its time window). Therefore, the execution time windows of both first subtasks include the virtual start time of the root node.

[0069] Step 203: If there is no conflict in resource requirements and the execution time window of the first subtask includes the virtual start time of the root node, then the first subtask is determined as a first-level child node directly under the root node.

[0070] Furthermore, when the resource requirements of the first subtask do not conflict with the initial resource pool of the root node, and its execution time window includes the virtual start time of the root node, it indicates that the first subtask meets the conditions to be a direct subordinate subtask of the root node. The warehouse inventory control system will officially determine the first subtask that meets the conditions as a first-level sub-node directly under the root node.

[0071] Continuing with the above embodiment, both subtask 1 and subtask 4 satisfy the conditions that their resource requirements are conflict-free and their execution time windows include the virtual start time of the root node. Therefore, the warehouse inventory control system determines subtask 1 and subtask 4 as first-level child nodes directly under the root node. At this time, the first layer of the initial DOM task structure tree is formed, with the two first-level child nodes, subtask 1 and subtask 4, directly connected under the root node.

[0072] Step 204: For any two nodes in the first-level child nodes, if the two nodes satisfy both resource compatibility and time compatibility, then the two nodes are determined to be parallel siblings, and all nodes with parallel sibling relationships are grouped into the same sibling group, thus obtaining the sibling relationship group of the first-level child nodes. Resource compatibility indicates that the resource requirements of the two nodes do not overlap, and time compatibility indicates that the execution time windows of the two nodes overlap.

[0073] Furthermore, for any two nodes within the first-level sub-nodes, the warehouse inventory control system assesses compatibility based on two dimensions: resource compatibility and time compatibility. Resource compatibility indicates that the resource requirements of the two nodes do not overlap; that is, the resources required by the two sub-tasks (such as forklifts, operators, specific equipment, etc.) do not overlap, and the same resource will not be required simultaneously. Time compatibility indicates that the execution time windows of the two nodes overlap; that is, the two sub-tasks have time intervals during which they can be executed simultaneously. When two first-level sub-nodes simultaneously satisfy both resource compatibility and time compatibility, the warehouse inventory control system determines that these two nodes are parallel siblings. Subsequently, the warehouse inventory control system groups all nodes with parallel sibling relationships into the same sibling group, obtaining the sibling relationship group of the first-level sub-nodes. Therefore, it can determine whether tasks can be executed in parallel during simulation.

[0074] Continuing with the above embodiment, the first-level sub-nodes include sub-task 1 (9:00-9:30 unloading material A, resource requirements: 1 forklift, 2 operators) and sub-task 4 (9:10-9:40 picking material B, resource requirements: 1 forklift, 1 operator):

[0075] Resource compatibility check: Subtask 1 requires 1 forklift and 2 operators, and Subtask 4 requires 1 forklift and 1 operator. The required resources do not overlap (the forklifts are different units, and the number and positions of operators do not conflict), therefore, resources are compatible. Time compatibility check: The time window for Subtask 1 is 9:00-9:30, and the time window for Subtask 4 is 9:10-9:40. The two time windows overlap between 9:10 and 9:30, therefore, time compatibility is achieved.

[0076] In summary, subtask 1 and subtask 4 are parallel siblings and are grouped into the same sibling group.

[0077] Step 205: Construct the initial DOM task structure tree based on the branch triggering dimension of the nodes in the sibling relationship group of the first-level child node.

[0078] Furthermore, the warehouse inventory control system constructs an initial DOM task structure tree based on the branch trigger dimension of the nodes in the sibling relationship group of the first-level child node, as detailed in steps 2051 to 2055.

[0079] This invention, through checks on resource requirements and time windows, ensures that initial resource contention and time misalignment do not occur during the execution of first-level subtasks. Matching based on sequential dependencies guarantees a reasonable order of task execution. First-level sub-nodes, as the starting operation of the overall task, conform to the overall task execution logic, making the entire task structure tree clearly present the task execution logic and facilitating the detection of conflicting tasks.

[0080] In one embodiment, steps 2051 to 2055 include:

[0081] Step 2051: For each sibling relationship group of the first-level child node, determine the nodes in the sibling relationship group whose precondition type is absolute precondition type and condition precondition type as candidate nodes.

[0082] Optionally, for each sibling relationship group of the first-level child node, the warehouse inventory control system analyzes the nodes within the group and filters them according to the pre-constraint type of the sub-task. In this embodiment of the invention, the pre-constraint types include absolute pre-constraint type and conditional pre-constraint type.

[0083] In this system, absolute prerequisite types mean that a subtask can only be executed after a specific prerequisite task is completed; this constraint is mandatory and unaffected by other factors. Conditional prerequisite types mean that the execution of a subtask depends on whether the execution result of a prerequisite task meets a specific condition; the subtask can only be executed if the condition is met. Therefore, the warehouse inventory control system identifies nodes with both types of prerequisite constraints as candidate nodes, and these candidate nodes have the potential to become secondary sub-nodes.

[0084] Continuing with the above embodiment, the sibling relationship group of the first-level child node includes subtask 1 (unloading 20 boxes of material A) and subtask 4 (picking 10 boxes of material B). Among the other subtasks corresponding to this sibling relationship group:

[0085] Subtask 2 (inspecting 20 boxes of material A) has an absolute prerequisite constraint (it must be executed after subtask 1 is completed); subtask 5 (packing 10 boxes of material B) has an absolute prerequisite constraint (it must be executed after subtask 4 is completed); assuming there is a subtask 8 (providing special reinforcement packaging for material A), its prerequisite constraint is a conditional prerequisite constraint (executed when the inspection results of subtask 2 show that material A is fragile). The warehouse inventory control system identifies the nodes corresponding to subtasks 2, 5, and 8 as candidate nodes.

[0086] Step 2052: For the first candidate node whose prerequisite constraint type is absolute, if the node corresponding to the prerequisite task pointed to by the subtask of the first candidate node is a first-level child node, then the first candidate node is determined as a second-level child node.

[0087] Furthermore, the warehouse inventory control system identifies the first candidate node (i.e., the candidate node with an absolute prerequisite constraint type), and then analyzes the node corresponding to the prerequisite task pointed to by the subtask of the first candidate node. If the prerequisite task of the subtask corresponding to the first candidate node is a subtask corresponding to a first-level sub-node, that is, the node corresponding to the prerequisite task is a first-level sub-node, then it indicates that the first candidate node has a direct absolute dependency relationship with the first-level sub-node, which meets the hierarchical positioning of a second-level sub-node, and therefore it is determined as a second-level sub-node.

[0088] Continuing with the candidate nodes identified above, subtask 2 (inspecting 20 boxes of material A) and subtask 5 (packing 10 boxes of material B) are the first candidate nodes with absolute prerequisite constraints. The prerequisite task for subtask 2 is subtask 1 (unloading 20 boxes of material A), and the node corresponding to subtask 1 is a first-level sub-node. Similarly, the prerequisite task for subtask 5 is subtask 4 (picking 10 boxes of material B), and the node corresponding to subtask 4 is also a first-level sub-node. Therefore, the warehouse inventory control system determines the nodes corresponding to subtask 2 and subtask 5 as second-level sub-nodes.

[0089] Step 2053: For the second candidate node with a precondition type of precondition, if the execution result of the subtask corresponding to the first-level child node satisfies the triggering condition of the subtask corresponding to the second candidate node, then the second candidate node is determined as a second-level child node.

[0090] Furthermore, the warehouse inventory control system focuses on the second candidate node (i.e., the candidate node with a conditional precondition type as its prerequisite constraint), analyzing the triggering conditions of the subtasks corresponding to the second candidate node, and the relationship between these triggering conditions and the execution results of the subtasks corresponding to the first-level sub-nodes. If the execution result of the subtask corresponding to the first-level sub-node satisfies the triggering condition of the subtask corresponding to the second candidate node after execution, it indicates that the conditional precondition type subtask has the prerequisite for execution, and the second candidate node is then determined as a second-level sub-node.

[0091] Continuing with the above embodiment, candidate node subtask 8 (special reinforcement packaging of material A) is the second candidate node with a pre-condition constraint type, and its trigger condition is "the quality inspection result of subtask 2 shows that material A is fragile". When subtask 2 (quality inspection of 20 boxes of material A) is completed, if the execution result shows that material A is fragile, the trigger condition of subtask 8 is met. At this time, the warehouse inventory control system determines the node corresponding to subtask 8 as a second-level sub-node.

[0092] Step 2054: For each second-level child node, bind the second-level child node to its corresponding first-level child node according to the sequential dependency dimension based on the time order constraint and resource inheritance constraint, to obtain the sequential dependency binding result. The time order constraint indicates that the end of the execution time window of the parent node is earlier than the start of the execution time window of the child node, and the resource inheritance constraint indicates that the resource release set of the parent node includes the initial resource requirements of the child node.

[0093] Furthermore, for each second-level child node, the warehouse inventory control system performs binding operations from two dimensions: time sequence constraints and resource acceptance constraints, to obtain the sequential dependency binding result. The sequential dependency binding result indicates whether the sequential dependency relationship between the second-level child node and the first-level child node is valid.

[0094] Optionally, in this embodiment of the invention, the time sequence constraint requires that the end of the execution time window of the parent node (i.e., the corresponding first-level child node) is earlier than the start of the execution time window of the child node (i.e., the second-level child node), ensuring that the child task can only start after the parent task is completed, thus avoiding time overlap and conflict. The resource assumption constraint requires that the resource release set of the parent node includes the initial resource requirements of the child node, that is, the resources released after the parent node finishes execution can meet the initial resources required when the child node starts execution, ensuring the smooth flow of resources.

[0095] Continuing with the above embodiment, for the second-level sub-node subtask 2 (inspecting 20 boxes of material A), its corresponding first-level sub-node is subtask 1 (unloading 20 boxes of material A): Time sequence constraint check: The execution time window end point of subtask 1 is 9:30, and the execution time window start point of subtask 2 is 9:20. At this time, the parent node's end point is earlier than the child node's start point, and the system records that this constraint is temporarily not satisfied. Resource acceptance constraint check: The resources released after the completion of subtask 1 include 1 forklift and 2 operators. The initial resource requirement of subtask 2 is 1 quality inspector. The resource release set of the parent node does not include the initial resource requirement of the child node, and the system records that this constraint is not satisfied.

[0096] Furthermore, the time window is adjusted so that the execution time window start point of subtask 2 is adjusted to 9:30, at which point the time sequence constraint is satisfied; at the same time, one quality inspector is assigned to subtask 2 to meet its initial resource requirements and resource acceptance constraints, and the final sequential dependency binding result is valid.

[0097] For the second-level sub-node subtask 5 (packing 10 boxes of material B), the corresponding first-level sub-node is subtask 4 (picking 10 boxes of material B). Time sequence constraint check: the execution time window end point of subtask 4 is 9:40, and the execution time window start point of subtask 5 is 9:30, which does not meet the constraint. Adjusting the start point of subtask 5 to 9:40 satisfies the constraint. Resource availability constraint check: subtask 4 releases 1 forklift and 1 operator, while subtask 5 initially requires 1 packing operator. After allocation, these are satisfied, and the sequential dependency binding result is valid. For the second-level sub-node subtask 8 (special reinforced packaging), the corresponding first-level sub-node is subtask 1: Time sequence constraint check: the end point of subtask 1 is 9:30, and the start point of subtask 8 is assumed to be 9:40, which satisfies the constraint. Resource availability constraint check: the forklift released by subtask 1 can be used for material handling in subtask 8, satisfying the initial resource requirements of subtask 8, and the sequential dependency binding result is valid.

[0098] Step 2055: Construct the initial DOM task structure tree based on the sequential dependency binding results of the second-level child nodes.

[0099] Furthermore, the warehouse inventory control system constructs an initial DOM task structure tree based on the sequential dependency binding results of the secondary child nodes, as detailed in steps 20551 to 20554.

[0100] This invention extends the task structure tree from the primary node to the secondary node by defining secondary child nodes, covering more sub-tasks and more comprehensively reflecting the entire material inbound / outbound task execution process. This ensures the structure tree includes not only primary tasks such as initial unloading and picking, but also secondary tasks such as quality inspection, packaging, and special packaging, presenting a complete task chain. By binding time sequence constraints and resource acceptance constraints, the sequential dependencies between secondary and primary child nodes are ensured to be real and effective, avoiding time conflicts and resource shortages.

[0101] In one embodiment, steps 20551 to 20554 include:

[0102] Step 20551: For all second-level child nodes under the same parent node, if the sequential dependency binding result between two second-level child nodes satisfies the sequential dependency binding condition, then construct the sequential dependency chain between the two second-level child nodes in the same branch according to the task order of the two second-level child nodes.

[0103] Optionally, the warehouse inventory control system focuses on all second-level child nodes under the same parent node (i.e., the same first-level child node), and analyzes these child nodes one by one. The sequential dependency binding condition is the key to determining whether two second-level child nodes can form a sequential dependency chain. This condition is based on the time order constraint and resource inheritance constraint in step 2054, that is, the end of the execution time window of the previous second-level child node is earlier than the start of the execution time window of the next second-level child node, and the resource release set of the previous second-level child node includes the initial resource requirements of the next second-level child node.

[0104] Furthermore, if two secondary child nodes satisfy the above-mentioned sequential dependency binding conditions, the warehouse inventory control system will construct a sequential dependency chain within the same branch according to their task execution logic, connecting the two nodes in sequence to ensure that the tasks are executed sequentially within the same branch.

[0105] In one embodiment, under the first-level sub-node sub-task 1 (unloading 20 boxes of material A), there are two second-level sub-nodes: sub-task 2 (quality inspection of 20 boxes of material A, execution time window 9:30-10:00, resource requirement 1 quality inspector, resource release set is 1 quality inspector) and sub-task 9 (moving the quality-inspected material A to the temporary storage area, execution time window 10:00-10:30, initial resource requirement 1 handling personnel and 1 forklift, although the quality inspector released by sub-task 2 is not directly required, it can be allocated from other areas after system coordination, and the end time of sub-task 2, 10:00, is earlier than the start time of sub-task 9, 10:00). The warehouse inventory control system determines that these two second-level sub-nodes meet the sequential dependency binding condition, and thus constructs a sequential dependency chain: sub-task 2 → sub-task 9, clarifying that sub-task 9 can only be executed after sub-task 2 is completed.

[0106] Step 20552: For all child nodes under different parent nodes, perform cross-validation on the parallel compatibility dimension for two child nodes at the same level under different branches to obtain the parallel compatibility dimension validation results. The cross-validation on the parallel compatibility dimension includes whether the resource requirements of two child nodes at the same level under different branches have no overlap, and whether the execution time windows of the two child nodes overlap.

[0107] Furthermore, the warehouse inventory control system identifies different branches, each starting from a different first-level sub-node. "Same level" refers to the same task level, such as both being second-level sub-nodes. Cross-validation for parallel compatibility is conducted from two aspects: resource requirements and execution time windows. First, it checks whether the resource requirements of two sub-nodes at the same level under different branches are non-overlapping, meaning their required resources (such as forklifts, operators, equipment, etc.) do not overlap, and there will be no resource contention. Second, it checks whether the execution time windows of the two sub-nodes overlap, meaning the two tasks have a time interval where they can be executed simultaneously. Therefore, the warehouse inventory control system performs cross-validation for parallel compatibility on two sub-nodes at the same level under different branches, obtaining the parallel compatibility verification result. This result will be used to determine whether to establish a cross-branch parallel sibling relationship.

[0108] Continuing with the above embodiment, different branches start with the first-level sub-nodes Subtask 1 (unloading 20 boxes of material A) and Subtask 4 (picking 10 boxes of material B), respectively. The second-level sub-nodes at the same level include Subtask 2 (quality inspection of 20 boxes of material A, resource requirement of 1 quality inspector, execution time window 9:30-10:00) and Subtask 5 (packing 10 boxes of material B, resource requirement of 1 packer, execution time window 9:40-10:10). The warehouse inventory control system performs cross-validation on these two sub-nodes: In terms of resource requirements, Subtask 2 requires a quality inspector, and Subtask 5 requires a packer; their resource requirements do not overlap. In terms of time windows, 9:40-10:00 is the overlapping period for both, indicating time overlap. Therefore, the parallel compatibility verification result shows that both resource compatibility and time compatibility are satisfied.

[0109] Step 20553: If the parallel compatibility dimension verification result satisfies resource compatibility and time compatibility, then establish a cross-branch parallel sibling relationship between two child nodes at the same level under different branches.

[0110] Furthermore, when the verification results of step 20552 show that two child nodes at the same level under different branches satisfy resource compatibility (no overlap in resource requirements) and time compatibility (overlapping execution time windows), it indicates that these two subtasks can be executed simultaneously without resource or time conflicts. Therefore, the warehouse inventory control system establishes a cross-branch parallel sibling relationship for these two child nodes, clarifying that they can be executed in parallel. Establishing this relationship helps improve the overall task execution efficiency and fully utilize resources and time.

[0111] Continuing with the above embodiment, the parallel compatibility verification results for subtask 2 (quality inspection of 20 boxes of material A) and subtask 5 (packing 10 boxes of material B) show that they meet both resource and time compatibility requirements. The warehouse inventory control system establishes a cross-branch parallel sibling relationship for these two sub-nodes, meaning that subtask 2 and subtask 5 can be executed simultaneously. For example, during the overlapping time period of 9:40-10:00, they can perform quality inspection and packaging operations respectively without interfering with each other.

[0112] Step 20554: Based on the sequential dependency chain of two second-level child nodes in the same branch in each level, the sibling relationship group in each level, and the cross-branch parallel sibling relationship between two child nodes in the same level under different branches in each level, establish the mutual connection relationship between all child nodes until all child nodes have corresponding connections, and obtain the initial DOM task structure tree.

[0113] Furthermore, the warehouse inventory control system summarizes the various relationships obtained in the previous steps: sequential dependency chains within the same branch (such as the chain constructed in step 20551), sibling relationship groups at each level (such as the sibling relationship group of first-level child nodes), and cross-branch parallel sibling relationships of child nodes at the same level under different branches (such as the relationship established in step 20553). Based on these relationships, the warehouse inventory control system establishes interconnections between all child nodes, ensuring that each child node has a corresponding connection with other related child nodes, whether it is a sequential dependency relationship, a sibling relationship within the same branch, or a cross-branch parallel relationship. Through this comprehensive connection, a complete and clear task structure tree is formed, which includes all subtasks, the attributes of subtasks, and various task relationships between subtasks.

[0114] Continuing with the above embodiments, all child nodes include:

[0115] First-level child nodes: Subtask 1, Subtask 4 (sibling group); Second-level child nodes: Subtask 2 (child node of Subtask 1), Subtask 5 (child node of Subtask 4), Subtask 8 (child node of Subtask 1), Subtask 9 (child node of Subtask 1, forming a sequential dependency chain with Subtask 2: Subtask 2 → Subtask 9);

[0116] Cross-branch parallel sibling relationship: Subtask 2 and Subtask 5.

[0117] Therefore, the warehouse inventory control system integrates these relationships and establishes connections between all child nodes: Subtask 1 is connected to Subtask 2 and Subtask 8 respectively; Subtask 4 is connected to Subtask 5; Subtask 2 is connected to Subtask 9; Subtask 2 and Subtask 5 establish a cross-branch parallel relationship; Subtask 1 and Subtask 4 are siblings. All child nodes have corresponding connections, ultimately forming the initial DOM task structure tree.

[0118] This invention, through constructing sequential dependency chains within the same branch, verifying cross-branch parallel compatibility, establishing cross-branch parallel relationships, and integrating all connection relationships, ultimately forms a complete and logically rigorous initial DOM task structure tree. In this initial DOM task structure tree, the establishment of cross-branch parallel sibling relationships allows tasks on different branches to be executed simultaneously when conditions permit, reducing the overall task execution time. All child nodes are connected through corresponding relationships, encompassing various relationships such as sequential dependency, sibling relationships within the same branch, and cross-branch parallelism, comprehensively reflecting the complex relationships between tasks and ensuring accurate representation of the task execution flow in subsequent task simulations.

[0119] In one embodiment, steps 301 to 303 include:

[0120] Step 301: Establish an association mapping between the nodes corresponding to all subtasks in the initial DOM task structure tree and the material inbound / outbound operation units in the warehouse digital twin model, thus obtaining the association mapping relationship. The association mapping rules are as follows: the resource requirements of the nodes corresponding to the subtasks correspond to the resource units in the warehouse digital twin model, and the execution time window of the nodes corresponding to the subtasks corresponds to the time axis in the warehouse digital twin model.

[0121] Optionally, an association mapping is established between the nodes corresponding to all subtasks in the initial DOM task structure tree and the material inbound / outbound operation units in the warehouse digital twin model. This association mapping relationship includes two aspects: First, the resource requirements of the nodes corresponding to the subtasks correspond to the resource units in the warehouse digital twin model. That is, the resources required by the subtasks, such as forklifts, operators, and shelf locations, must be matched with corresponding entity resource units in the digital twin model. For example, if a subtask requires one forklift, it corresponds to a forklift resource unit with a specific number in the model. Second, the execution time window of the nodes corresponding to the subtasks corresponds to the time axis in the warehouse digital twin model. The start and end times of the subtasks must correspond to specific times on the time axis in the model. For example, if the execution time window of a subtask is 9:00-9:30, it corresponds to the time period from 9:00 to 9:30 on the model's time axis.

[0122] Through the aforementioned association mapping, the virtual subtask nodes are linked to the actual operational units and time dimension in the digital twin model, ensuring the realism and accuracy of the simulation process.

[0123] In one embodiment, the subtasks in the initial DOM task structure tree include:

[0124] Subtask 1 (Unloading 20 boxes of material A): Resource requirements: 1 forklift (F01), 2 operators (P01, P02), execution time window: 9:00-9:30; Subtask 2 (Inspecting 20 boxes of material A): Resource requirements: 1 quality inspector (Q01), execution time window: 9:30-10:00; Subtask 4 (Picking 10 boxes of material B): Resource requirements: 1 forklift (F02), 1 operator (P03), execution time window: 9:10-9:40; Subtask 5 (Packaging 10 boxes of material B): Resource requirements: 1 packer (P04), execution time window: 9:40-10:10.

[0125] The warehouse inventory control system performs association mapping according to association mapping rules:

[0126] The resource requirements for subtask 1 correspond to forklift F01, operators P01 and P02 in the warehouse digital twin model, with an execution time window of 9:00-9:30 corresponding to 9:00-9:30 on the model's timeline; the resource requirements for subtask 2 correspond to quality inspector Q01 in the model, with an execution time window of 9:30-10:00 corresponding to 9:30-10:00 on the model's timeline; the resource requirements for subtask 4 correspond to forklift F02 and operator P03 in the model, with an execution time window of 9:10-9:40 corresponding to 9:10-9:40 on the model's timeline; the resource requirements for subtask 5 correspond to packer P04 in the model, with an execution time window of 9:40-10:10 corresponding to 9:40-10:10 on the model's timeline.

[0127] This yields the association mapping relationship between each subtask node and the digital twin model operation unit.

[0128] Step 302: Based on the association mapping relationship and the sequential dependency chain under the same branch in the initial DOM task structure tree, the execution simulation of subtasks is triggered sequentially according to the sequential dependency relationship in the warehouse digital twin model to obtain the execution simulation trigger sequence. The triggering rule is: when the execution simulation status of the current subtask is completed, the execution simulation of the subsequent subtask is triggered.

[0129] Furthermore, the warehouse inventory control system performs simulations based on the associated mapping relationships and the sequential dependency chains within the same branch of the initial DOM task structure tree. The triggering rule is that the execution simulation of subsequent subtasks can only be triggered when the current subtask's execution simulation status is complete. The system first starts the execution simulation of the initial subtask in the sequential dependency chain within the digital twin model, monitoring its execution simulation status in real time. Once the initial subtask's simulation status changes to complete, the execution simulation of the next subtask is immediately triggered according to the sequential dependency chain, and so on, until all subtasks on the sequential dependency chain have completed their simulations, forming an execution simulation trigger sequence. This sequential triggering method ensures that the execution simulations of subtasks within the same branch strictly follow the order, conforming to the logic of task execution.

[0130] Continuing with the above embodiment, the initial DOM task structure tree contains sequential dependency chains under the same branch: Subtask 1 → Subtask 2 → Subtask 9 (moving material A to the temporary storage area). The warehouse inventory control system simulates and triggers based on the associated mapping relationship and triggering rules: the execution simulation of subtask 1 is started in the warehouse digital twin model, with the simulation time starting at 9:00. At 9:30, the execution simulation status of subtask 1 changes to complete. Since the status of subtask 1 is complete, the execution simulation of subtask 2 is triggered, with the simulation starting at 9:30. At 10:00, the execution simulation status of subtask 2 changes to complete. When the status of subtask 2 is complete, the execution simulation of subtask 9 is triggered, with the simulation starting at 10:00. Thus, the execution simulation triggering sequence is: Subtask 1 (triggered at 9:00) → Subtask 2 (triggered at 9:30) → Subtask 9 (triggered at 10:00).

[0131] Step 303: Determine the conflicting task based on the association mapping relationship and the execution of the simulated trigger sequence.

[0132] Furthermore, the warehouse inventory control system determines conflicting tasks based on the associated mapping relationship and the execution simulation trigger sequence, as detailed in steps 3031 to 3034.

[0133] This invention establishes an association mapping relationship, ensuring a close correspondence between subtasks and the resource and time dimensions in the digital twin model. Sequentially triggered simulations follow the task execution logic, guaranteeing a high degree of similarity between the simulation and actual execution processes, thus laying the foundation for accurate conflict detection. Conflict simulation comprehensively identifies various conflicting tasks, such as those involving time overlap, parallel sibling tasks, and resource consumption, ensuring the targeted and effective optimization of subsequent tasks. This makes the material inbound and outbound tasks in VMI physical warehouses run more smoothly and efficiently, achieving effective simulation of the material inbound and outbound task execution process within the warehouse digital twin model and accurately identifying conflicting tasks with anomalies.

[0134] In one embodiment, steps 3031 to 3034 include:

[0135] Step 3031: Based on the execution simulation trigger sequence, identify time-sequence conflicts in the sequential dependency chain of tasks in the warehouse digital twin model to obtain time-sequence conflict task pairs. The identification rule is: if the execution simulation end time of the preceding subtask in the sequential dependency chain is later than the execution simulation start time of the following subtask, it is marked as a time-sequence conflict task pair.

[0136] Optionally, the warehouse inventory control system, based on the execution simulation trigger sequence, focuses on the sequential dependency chain under the same branch in the initial DOM task structure tree, and monitors the execution simulation status of subtasks in the chain. Optionally, the identification rule in this embodiment of the invention is: if the execution simulation end time of the preceding subtask in the sequential dependency chain is later than the execution simulation start time of the following subtask, then these two subtasks constitute a time-sequential conflict task pair.

[0137] Therefore, the warehouse inventory control system will check each pair of preceding and following subtasks in the sequential dependency chain one by one, compare their execution simulation end time and start time, and thus determine whether there is a time connection conflict.

[0138] Continuing with the above embodiment, in the sequential dependency chain subtask 2 (inspecting 20 boxes of material A) → subtask 9 (moving material A to the temporary storage area), the original planned end time for the simulation of subtask 2 was 10:00, and the original planned start time for the simulation of subtask 9 was 10:00. However, during the simulation, subtask 2 discovered some abnormalities in the packaging of material A, requiring additional time for inspection, thus delaying the end time of the simulation to 10:15. Subtask 9, however, still started its simulation at 10:00 as originally planned.

[0139] Based on the identification rules, the warehouse inventory control system determines that there is a time connection conflict in the execution simulation of subtask 2 and subtask 9, and marks them as a time connection conflict task pair.

[0140] Step 3032: Based on the association mapping relationship and the parallel sibling tasks belonging to the same sibling group in the initial DOM task structure tree, synchronous execution simulation monitoring is performed in the warehouse digital twin model to obtain conflicting parallel sibling task pairs. The monitoring content is: if the resource units corresponding to the parallel sibling tasks are occupied at the same time, and the execution time windows of the parallel sibling tasks have resource competition within overlapping intervals on the time axis of the warehouse digital twin model, then they are marked as conflicting parallel sibling task pairs.

[0141] Furthermore, for parallel sibling tasks belonging to the same sibling group in the initial DOM task structure tree, the warehouse inventory control system clarifies the resource units and execution time windows corresponding to the parallel sibling tasks based on the association mapping relationship. Based on the resource units and execution time windows corresponding to the parallel sibling tasks, the system performs synchronous execution simulation monitoring in the warehouse digital twin model. The monitoring content includes two aspects: first, checking whether the resource units corresponding to the parallel sibling tasks are occupied at the same time, that is, whether multiple parallel tasks use the same resource at the same time; second, checking whether there is resource competition within the overlapping interval of the execution time windows of the parallel sibling tasks, that is, whether the number of resources cannot meet the simultaneous needs of multiple parallel tasks.

[0142] If both conditions are met, the warehouse inventory control system will mark the pair of parallel sibling tasks as conflicting parallel sibling task pairs, thereby discovering the conflict in resource usage between parallel tasks.

[0143] Continuing with the above embodiment, the parallel sibling tasks within the same sibling group include subtask 1 (unloading 20 boxes of material A, requiring 1 forklift F01, execution time window 9:00-9:30) and subtask 4 (picking 10 boxes of material B, requiring 1 forklift F02, execution time window 9:10-9:40). During synchronous execution simulation monitoring, it was found that forklifts F01 and F02 need to pass through the same narrow passage between 9:20 and 9:30. This passage can only accommodate one forklift at a time, meaning both forklifts need to use the passage simultaneously, resulting in resource contention. Therefore, based on the monitoring data, the warehouse inventory control system determines that subtask 1 and subtask 4 constitute a conflicting parallel sibling task pair and marks them as such.

[0144] Step 3033: Based on the association mapping relationship and the execution simulation trigger sequence, combined with the cross-branch parallel tasks in the initial DOM task structure tree that have passed cross-branch parallel compatibility verification, resource occupation conflicts are detected in the warehouse digital twin model to obtain resource occupation conflict task pairs. The detection rule is: if a cross-branch parallel task is in the triggered execution state in the task execution simulation trigger sequence, and its corresponding resource unit is simultaneously requested and occupied during the execution simulation, then it is marked as a resource occupation conflict task pair.

[0145] Furthermore, for cross-branch parallel tasks in the initial DOM task structure tree that have undergone cross-branch parallel compatibility verification, the warehouse inventory control system determines the execution simulation state and corresponding resource units of the cross-branch parallel tasks based on the association mapping relationship and the execution simulation trigger sequence. It then detects resource occupancy conflicts in the warehouse digital twin model based on the execution simulation state and corresponding resource units of the cross-branch parallel tasks. The detection rule is: if a cross-branch parallel task is in the triggered execution state in the execution simulation trigger sequence, and its corresponding resource unit is simultaneously requested and occupied during the execution simulation, then these two cross-branch parallel tasks constitute a resource occupancy conflict task pair. Therefore, by monitoring the resource requests and occupancy of cross-branch parallel tasks in real time, the warehouse inventory control system promptly detects resource occupancy conflicts, ensuring that parallel tasks in different branches do not conflict in resource usage.

[0146] Continuing with the above embodiment, the cross-branch parallel tasks that have passed cross-branch parallel compatibility verification include subtask 2 (inspecting 20 boxes of material A, requiring 1 quality inspector Q01, execution time window 9:30-10:00) and subtask 5 (packing 10 boxes of material B, requiring 1 operator P04, execution time window 9:40-10:10). During the simulation, subtask 2 temporarily requires operator P04 to assist in moving quality inspection tools due to an increase in the quality inspection workload. At the same time, subtask 5 is using P04 for packing operations, resulting in P04 being simultaneously requested for use by subtask 2 and subtask 5 from 9:50 to 10:00. Therefore, the warehouse inventory control system, according to the detection rules, determines that subtask 2 and subtask 5 constitute a resource-occupation conflict task pair and marks them as such.

[0147] Step 3034: Identify time-coordinated conflict task pairs, conflicting parallel sibling task pairs, and resource-occupation conflict task pairs as conflicting tasks.

[0148] Furthermore, the warehouse inventory control system integrates time-series conflict task pairs, conflicting parallel sibling task pairs, and resource-occupation conflict task pairs. These task pairs together constitute the conflicting tasks that are abnormal during the simulation process.

[0149] Continuing with the above embodiment, step 3031 identifies time-coordinated conflicting task pairs: subtask 2 and subtask 9; step 3032 identifies conflicting parallel sibling task pairs: subtask 1 and subtask 4; step 3033 identifies resource-occupancy conflicting task pairs: subtask 2 and subtask 5. The warehouse inventory control system aggregates these task pairs and determines that conflicting tasks include subtask 2 and subtask 9, subtask 1 and subtask 4, and subtask 2 and subtask 5.

[0150] This invention identifies time-sequence conflicts, parallel sibling conflicts, and resource-occupancy conflicts by categorizing them. This comprehensive and accurate identification of all conflicting tasks exhibiting anomalies during the simulation process ensures comprehensive coverage. On one hand, it encompasses time conflicts within sequential dependency chains, parallel resource conflicts within the same sibling group, and resource conflicts between parallel tasks across branches, guaranteeing that all possible conflict types are included in the monitoring scope without overlooking any important conflicting tasks. On the other hand, it clarifies the identification rules and monitoring content for each pair of conflicting tasks, making judgments based on real-time simulation data from the warehouse digital twin model. This ensures that the identified conflicting tasks accurately reflect potential problems during task execution. The accurate identification of conflicting tasks provides a reliable basis for subsequent task optimization, facilitating the development of targeted optimization strategies and improving the execution efficiency and stability of VMI physical warehouse material inbound and outbound tasks.

[0151] In one embodiment, the process from steps A to D includes:

[0152] Step A: Define the scope of influence of conflicting tasks based on their associated attributes.

[0153] Optionally, the warehouse inventory control system determines the associated attributes of conflicting tasks, including the intersection of resource requirements of the conflicting tasks (i.e., the resources required by the two conflicting tasks), the intersection of execution time windows (i.e., the intervals in which the execution times of the two conflicting tasks overlap), the branch information (i.e., which branch each conflicting task belongs to), and the node level in the initial DOM task structure tree (i.e., the hierarchical position of the conflicting task in the structure tree).

[0154] Furthermore, the warehouse inventory control system determines the scope of impact based on the relationship and task type of the conflicting tasks, according to the following definition rules: If the conflicting tasks belong to the same sequential dependency chain, it means they have a sequential execution relationship on the same branch. The scope of impact includes all subsequent subtasks in that sequential dependency chain (because conflicts in preceding tasks will affect the execution of subsequent tasks), as well as related subtasks in the execution simulation trigger sequence that are related to that sequential dependency chain (these related subtasks may depend on tasks in that chain). If the conflicting tasks belong to parallel tasks on different branches, it means they are executed simultaneously on different branches. The scope of impact includes all subtasks in the two branches that share resources (i.e., use the same resources) or overlap in time (i.e., their execution times intersect), as well as related subtasks in the execution simulation trigger sequence that are related to these two branches (these related subtasks may interact with tasks in the two branches).

[0155] Continuing with the above embodiment, the conflicting tasks belong to the same sequential dependency chain for subtasks 2 and 9 (subtask 1 → subtask 2 → subtask 9). The warehouse inventory control system analyzes their associated attributes: the resource requirement intersection is empty (subtask 2 requires quality inspectors, subtask 9 requires handling personnel and forklifts), the execution time window intersection is 10:00-10:10 (subtask 2's end time is delayed to 10:10, subtask 9's start time is 10:00), the branch information is the branch where subtask 1 is located, and the node level is a second-level sub-node.

[0156] According to the definition rules, its scope of influence includes subsequent subtasks in the sequential dependency chain (if there are no subsequent subtasks), as well as associated subtasks related to the chain in the execution simulation trigger sequence (such as subtask 1, because subtask 1 is the starting task of the chain).

[0157] The conflicting tasks are parallel tasks belonging to different branches of subtasks 2 and 5 (subtask 2 belongs to the branch of subtask 1, and subtask 5 belongs to the branch of subtask 4). Related attributes: the resource requirement intersection is operator P04, the execution time window intersection is 9:50-10:00, the branch information is the branch of subtask 1 and the branch of subtask 4 respectively, and the node level is a second-level child node.

[0158] According to the definition rules, its scope of influence includes subtasks in the two branches that share resources (using P04) or overlap in time (9:50-10:00) with them (such as subtask 9 in the branch where subtask 1 is located, and subtask 6 in the branch where subtask 4 is located), as well as related subtasks in the execution simulation trigger sequence that are related to the two branches (such as subtask 1 and subtask 4).

[0159] Step B: If the intersection of the resource requirements of the conflicting tasks is non-empty, and the resource requirements of all affected subtasks within the scope of influence are contained within the intersection of resource requirements, then the task type of the conflicting task is determined to be a resource allocation conflict type.

[0160] Furthermore, the warehouse inventory control system checks whether the intersection of resource requirements for conflicting tasks is non-empty, meaning whether the two conflicting tasks have resources they both need. If the intersection of resource requirements is non-empty, the system further analyzes the relationship between the resource requirements of all affected sub-tasks within the affected area and this intersection. If the resource requirements of all affected sub-tasks within the affected area have an inclusion relationship with the intersection of resource requirements (i.e., the resource requirements of the affected sub-tasks all contain the resources commonly needed by the conflicting tasks), it indicates that these sub-tasks all depend on the conflicting resources, and the root cause of the conflict lies in unreasonable resource allocation. Therefore, the task type of the conflicting tasks is determined to be a resource allocation conflict type.

[0161] Continuing with the above embodiment, the conflicting tasks for subtasks 2 and 5 have the same resource requirement: operator P04 (not empty). The affected subtasks within the scope of influence include subtasks 1, 4, 9, and 6.

[0162] The resource requirements for subtask 1 are forklift F01 and operators P01 and P02, which have no inclusion relationship with the resource requirement intersection (P04); the resource requirements for subtask 4 are forklift F02 and operators P03, which have no inclusion relationship with the resource requirement intersection (P04); the resource requirements for subtask 9 are handling personnel and forklift, which have no inclusion relationship with the resource requirement intersection (P04); the resource requirement for subtask 6 is operator P04, which has an inclusion relationship with the resource requirement intersection (P04).

[0163] Since not all affected subtasks have an inclusion relationship with the resource requirement intersection, subtasks 2 and 5 do not belong to the resource allocation conflict type.

[0164] Taking another conflicting task pair as an example, suppose that the resource requirements of the conflicting task pair subtasks X and Y intersect at forklift F03 (not empty), and the resource requirements of the affected subtasks Z and W within the scope of influence both include forklift F03. Then the task type of subtasks X and Y is resource allocation conflict type.

[0165] Step C: If the intersection of the execution time windows of the conflicting tasks is non-empty, and the execution time windows of all affected subtasks within the scope of influence overlap with the intersection of their execution time windows, and the execution simulation end time of the preceding subtask is later than the execution simulation start time of the following subtask, then the task type of the conflicting task is determined to be a time planning conflict type.

[0166] Furthermore, the warehouse inventory control system checks whether the intersection of the execution time windows of the conflicting tasks is non-empty, that is, whether the execution times of the two conflicting tasks overlap. If the intersection of the execution time windows is non-empty, the system then analyzes the relationship between the execution time windows of all affected subtasks within the affected area and this intersection. If the execution time windows of all affected subtasks within the affected area overlap with the intersection of their execution time windows, and there is a situation where the execution simulation end time of the preceding subtask is later than the execution simulation start time of the following subtask, then it indicates that the root cause of the conflict lies in unreasonable time planning, and therefore the task type of the conflicting tasks is determined to be a time planning conflict type.

[0167] Continuing with the above embodiment, for the conflicting tasks, subtasks 2 and 9 have an intersection of execution time windows of 10:00-10:10 (not empty). The affected subtask within the scope of influence is subtask 1. Subtask 1's execution time window is 9:00-9:30, which does not overlap with the intersection of 10:00-10:10. Therefore, subtasks 2 and 9 do not belong to the time planning conflict type.

[0168] In another example, the execution time windows of conflicting tasks M and N intersect at 14:00-14:30 (not empty). The execution time windows of affected subtasks O and P within the scope of influence also overlap with 14:00-14:30. Furthermore, the execution simulation end time of subtask M (preceding task) at 14:20 is later than the execution simulation start time of subtask N (following task) at 14:10. Therefore, the task type of subtasks M and N is a time planning conflict type.

[0169] Step D: If the branch information of the conflicting task does not match, and all affected subtasks within the scope of influence belong to two different branches, and the cross-branch parallel compatibility results do not meet resource compatibility and / or time compatibility, then the task type of the conflicting task is determined to be a branch coordination conflict type.

[0170] Furthermore, the warehouse inventory control system checks whether the branch information of the conflicting tasks matches, i.e., whether the two conflicting tasks belong to the same branch. If the branch information does not match (i.e., they belong to different branches), the system analyzes the branch affiliation of all affected subtasks within the scope of influence. If all affected subtasks within the scope of influence belong to two different branches, and the cross-branch parallel compatibility results do not meet resource compatibility (intersecting resource requirements) and / or time compatibility (no overlapping time windows), then the root cause of the conflict lies in improper coordination between different branches, and the task type of the conflicting task is determined to be a branch coordination conflict type.

[0171] Continuing with the above embodiment, the conflicting tasks for subtasks 2 and 5 are as follows: their branch information does not match (subtask 2 belongs to the branch of subtask 1, and subtask 5 belongs to the branch of subtask 4). The affected subtasks within the scope of influence include subtask 1 (the branch of subtask 1), subtask 4 (the branch of subtask 4), subtask 9 (the branch of subtask 1), and subtask 6 (the branch of subtask 4), meaning all affected subtasks belong to two different branches. In the cross-branch parallel compatibility result, the resource requirements of both subtasks intersect at operator P04 (not satisfying resource compatibility), and their time windows overlap (satisfying time compatibility), meaning the cross-branch parallel compatibility result does not satisfy resource compatibility. Therefore, the task type for subtasks 2 and 5 is a branch coordination conflict type.

[0172] This invention combines the associated attributes and scope of influence of conflicting tasks for comprehensive analysis, ensuring accurate judgment of task types and laying a solid foundation for developing targeted optimization strategies. This helps improve the execution efficiency and stability of VMI physical warehouse material inbound and outbound tasks.

[0173] Furthermore, for tasks with resource allocation conflicts, the process of optimizing the execution strategy based on the task nodes and task types of the conflicting tasks includes:

[0174] Step 401: Decompose the intersection of resource requirements of the first objective task with resource allocation conflicts into multiple subsets of minimum resource units. Each subset of minimum resource units corresponds to each independent resource type in the warehouse digital twin model.

[0175] Optionally, the warehouse inventory control system identifies the primary target task (i.e., the task with resource allocation conflicts) and determines the intersection of their resource requirements, that is, the set of resources commonly needed by both tasks. Then, the warehouse inventory control system decomposes this resource requirement intersection according to resource type, obtaining multiple subsets of minimum resource units. Each subset of minimum resource units corresponds to an independent resource type in the warehouse digital twin model, such as forklifts, operators, or specific equipment, ensuring that the resource units in each subset belong to the same type and cannot be further divided.

[0176] In one embodiment, the first target task is sub-task X (handling material A) and sub-task Y (handling material B). There is a resource allocation conflict between the two, and the resource requirements overlap as "2 forklifts (F01, F02) and 1 operator (P01)".

[0177] The warehouse inventory control system decomposes the intersection of resource demands into subsets of minimum resource units: Minimum resource unit subset 1: {Forklift F01, Forklift F02} (corresponding to forklift resource types); Minimum resource unit subset 2: {Operator P01} (corresponding to operator resource type). Each subset corresponds to an independent resource type in the digital twin model.

[0178] Step 402: Divide the resource pools of all affected subtasks within the scope of the first target task into dedicated resource pools corresponding to the smallest resource unit subset, thus obtaining the branch-specific resource pools.

[0179] Furthermore, for each branch to which all affected subtasks within the scope of the first target task belong, the warehouse inventory control system divides the resource pool of each branch, making it correspond to the minimum resource unit subset in step 401, forming a branch-specific resource pool. Each branch-specific resource pool not only includes the corresponding minimum resource unit subset, but also other resource units in that branch that are not conflict-occupied, ensuring that each branch has independent resource supply and avoiding resource competition between different branches. Continuing with the above embodiment, subtask X of the first target task belongs to branch 1, subtask Y belongs to branch 2, and all affected subtasks within the scope of influence belong to these two branches. Dividing branch-specific resource pools:

[0180] Branch 1's dedicated resource pool includes forklift F01 from the smallest resource unit subset 1, forklift F03 from branch 1 that is not conflicted, operator P02, and other resources outside of the smallest resource unit subset 2. Branch 2's dedicated resource pool includes forklift F02 from the smallest resource unit subset 1, forklift F04 from branch 2 that is not conflicted, operator P03, and other resources outside of the smallest resource unit subset 2. This way, the two branches have independent resource pools, avoiding resource conflicts.

[0181] Step 403: Based on the branch-specific resource pool and the resource requirement priority of the first target task, allocate a corresponding target subset to the first target task. The resource requirement priority is determined based on its position in the sequential dependency chain.

[0182] Furthermore, the warehouse inventory control system allocates resources based on the branch-specific resource pool and the resource requirement priority of the primary objective task. The resource requirement priority is determined by the task's position in the sequential dependency chain; generally, tasks earlier in the chain have higher priority because their execution results affect subsequent tasks.

[0183] Furthermore, the warehouse inventory control system allocates a corresponding target subset (i.e., the smallest resource unit subset that satisfies its resource requirements) from the corresponding branch-specific resource pool for each primary target task according to priority.

[0184] Continuing with the branch-specific resource pools described above, subtask X has a higher priority in the dependency chain of branch 1, while subtask Y has a lower priority in the dependency chain of branch 2. Assign target subsets to the first target task: assign forklift F01 (from the smallest resource unit subset 1) from the branch 1 resource pool to subtask X (branch 1); assign forklift F02 (from the smallest resource unit subset 1) from the branch 2 resource pool to subtask Y (branch 2). Ensure that high-priority tasks receive the resources they need first.

[0185] Step 404: Update the resource requirements of all affected subtasks within the scope of influence of the first target task based on the target subset to obtain the adjusted resource requirements.

[0186] Furthermore, the warehouse inventory control system updates the resource requirements of all affected subtasks within the scope of the first target task based on the target subset, resulting in adjusted resource requirements. Specifically, the resource requirements of the affected subtasks that are related to conflicts (i.e., resources that originally belonged to the intersection of resource requirements) are replaced with resource units from the dedicated resource pool corresponding to their respective branches, ensuring that these subtasks no longer depend on conflicting resources and avoiding resource contention.

[0187] Continuing with the above embodiment, the affected subtasks within the scope of the first target task include subtask X1 (packaging material A) of branch 1 and subtask Y1 (packaging material B) of branch 2. Both of their original resource requirements included operator P01 (belonging to the intersection of resource requirements). Updated resource requirements: Adjusted resource requirements for subtask X1: Replace operator P01 with operator P02 from the branch 1 dedicated resource pool; Adjusted resource requirements for subtask Y1: Replace operator P01 with operator P03 from the branch 2 dedicated resource pool. After the update, the two subtasks no longer depend on the conflicting resource P01.

[0188] Step 405: Associate the intersection of the branch-specific resource pool and the execution time window of the first target task to obtain the association result.

[0189] Furthermore, the warehouse inventory control system correlates the intersection of each branch's dedicated resource pool with the execution time window of the primary target task to obtain the correlation result. Through this correlation, non-overlapping time periods are allocated to each branch's dedicated resource pool within the execution time window, ensuring that the same branch's dedicated resource pool is occupied by only one task in different time periods, avoiding the use of resources in the same resource pool by multiple tasks at the same time, and further eliminating resource conflicts.

[0190] Continuing with the above embodiment, the execution time windows of the first target task subtask X and subtask Y intersect at 8:00-9:00. Associating branch-specific resource pools with time windows: Allocating a time period of 8:00-8:30 to the branch 1 resource pool, during which only subtask X and its affected subtasks within this time period can use this resource pool; allocating a time period of 8:30-9:00 to the branch 2 resource pool, during which only subtask Y and its affected subtasks within this time period can use this resource pool. This ensures that the same resource pool is used in an orderly manner across different time periods.

[0191] Step 406: Based on the association results and the adjusted resource requirements, reconstruct the sequential dependency chain in the execution simulation trigger sequence of the first target task to obtain the optimized execution strategy.

[0192] Furthermore, the warehouse inventory control system reconstructs the sequential dependency chain in the execution simulation trigger sequence of the first target task based on the correlation results and adjusted resource requirements. During the reconstruction process, the triggering relationship of tasks in the sequential dependency chain is modified to a triggering relationship based on the adjusted resource requirements and correlation results. That is, the triggering of a task depends not only on the completion of the preceding task, but also on whether the adjusted resources are obtained and whether it is within the allocated time period. The resulting optimized execution strategy ensures that tasks are executed sequentially without resource conflicts.

[0193] Continuing with the above embodiment, the original sequential dependency chain is: Subtask X → Subtask X1; Subtask Y → Subtask Y1. The sequential dependency chain is reconstructed: Subtask X uses resources from the branch 1 dedicated resource pool from 8:00 to 8:30, and upon completion, triggers Subtask X1 (using the adjusted resource P02, executing within 8:00 to 8:30); Subtask Y uses resources from the branch 2 dedicated resource pool from 8:30 to 9:00, and upon completion, triggers Subtask Y1 (using the adjusted resource P03, executing within 8:30 to 9:00), forming an optimized execution strategy and eliminating resource conflicts.

[0194] This invention, through the decomposition of resource requirements into the smallest units and the creation of dedicated resource pools, achieves refined resource management and branch isolation, avoiding resource contention between different branches and improving resource utilization efficiency. The priority-based resource allocation and time isolation mechanism ensures that tasks are executed in a reasonable order and at appropriate times, eliminating interference from resource allocation conflicts and enabling the smooth progress of the primary objective task and its subtasks within its scope. Therefore, it provides a systematic optimization method for tasks with resource allocation conflicts, ensuring the coordinated and orderly execution of material inbound and outbound tasks in VMI physical warehouses at the resource level, and providing strong support for the efficient execution of the overall tasks.

[0195] Furthermore, for tasks with time-planning conflicts, the process of optimizing the execution strategy based on the task nodes and task types of the conflicting tasks includes:

[0196] Step 407: For the second objective task with time planning conflicts, identify the time connection conflict interval between the execution simulation end time of the preceding subtask and the execution simulation start time of the subsequent subtask based on the second objective task.

[0197] Optionally, for a second objective task with a time-planning conflict, the warehouse inventory control system identifies the preceding and succeeding subtasks that constitute the conflict, and then extracts the execution simulation end time of the preceding subtask and the execution simulation start time of the succeeding subtask. The time-series conflict interval is defined as the time period from the execution simulation end time of the preceding subtask to the execution simulation start time of the succeeding subtask. If the execution simulation end time of the preceding subtask is later than the execution simulation start time of the succeeding subtask, this interval represents the overlapping and conflicting portion of their times.

[0198] In one embodiment, the second target task consists of subtask M (preceding subtask, material inspection) and subtask N (subsequent subtask, material warehousing). The simulation end time for subtask M is 14:30, and the simulation start time for subtask N is 14:00. The warehouse inventory control system identifies the time overlap conflict interval as 14:00-14:30, which is the time period from the planned start time of subtask N to the actual end time of subtask M. Within this interval, the two tasks have a time overlap conflict.

[0199] Step 408: Based on the scope of influence of the second target task, determine the overlap between the execution time windows and time connection conflict intervals of all affected subtasks within the scope of influence, and obtain the overlap relationship judgment result.

[0200] Furthermore, the warehouse inventory control system determines the scope of influence of the second target task (i.e., all sub-tasks that may be affected by time conflicts), and then analyzes the intersection of the execution time windows and time-connection conflict intervals of these sub-tasks one by one. In this embodiment of the invention, the formula for calculating the overlap is: Overlap = Length of the intersection of the execution time windows of the affected sub-tasks and the time-connection conflict intervals ÷ Length of the execution time windows of the affected sub-tasks. By calculating the overlap, the result of the overlap relationship judgment is obtained (e.g., complete inclusion, partial overlap, no overlap).

[0201] Continuing with the above embodiment, within the influence range of the second target task (sub-tasks M and N), there are two affected sub-tasks: Sub-task O (material identification printing): its execution time window is 13:50-14:20, and its intersection with the time conflict interval (14:00-14:30) is 14:00-14:20 (length 20 minutes). Its own time window length is 30 minutes, and the overlap degree = 20 ÷ 30 ≈ 0.67 (partial overlap); Sub-task P (warehouse location planning): its execution time window is 14:10-14:40, and its intersection with the time conflict interval is 14:10-14:30 (length 20 minutes). Its own time window length is 30 minutes, and the overlap degree = 20 ÷ 30 ≈ 0.67 (partial overlap). The warehouse inventory control system obtains the following overlap relationship judgment result: both sub-tasks O and P partially overlap with the time conflict interval.

[0202] Step 409: Based on the overlap relationship judgment result, the execution time window of all affected subtasks in the influence range of the second target task is offset and adjusted to obtain the offset execution time window.

[0203] Furthermore, based on the overlap relationship judgment results, the warehouse inventory control system adjusts the execution time window of the affected subtasks to eliminate overlap with the time-connection conflict interval. The warehouse inventory control system adopts different adjustment strategies for different overlap relationships: if the execution time window of the affected subtask completely includes the time-connection conflict interval (overlap degree = 1), it is shifted backward as a whole until the time-connection conflict interval ends, ensuring that the entire time window avoids the conflict interval; if there is only partial overlap (0 < overlap degree < 1), the overlapping part is cut and shifted backward, while the non-overlapping part retains its original time, ensuring that the execution time window after the shift does not overlap with the conflict interval, and the whole is still within the reasonable expansion range of the original time window.

[0204] Continuing with the above embodiment, adjustments are made to the affected subtasks: Subtask O (13:50-14:20): The overlapping portion with the conflict interval (14:00-14:30) is 14:00-14:20. This portion is shifted backward to 14:30-14:50. After the adjustment, the execution time window is 13:50-14:00 (non-overlapping portion) and 14:30-14:50 (shifted portion); Subtask P (14:10-14:40): The overlapping portion with the conflict interval is 14:10-14:30. This portion is shifted backward to 14:30-14:50. After the adjustment, the execution time window is 14:30-14:50 (overall shift; since the non-overlapping portion is only 14:30-14:40, it is segmented and merged into a continuous window). After the adjustment, the time windows of both subtasks do not overlap with the conflict interval.

[0205] Step 410: Based on the time connection conflict interval, compress the execution flow of the preceding subtask to obtain the compressed execution flow of the preceding subtask, and split the execution flow of the following subtask based on the time connection conflict interval to obtain two sub-flows of the split following subtask.

[0206] Furthermore, based on the time-series conflict interval, the execution flow of the preceding subtask is compressed to obtain a compressed preceding subtask execution flow. Then, based on the time-series conflict interval, the execution flow of the subsequent subtask is split, resulting in two sub-flows for the split subsequent subtask. The compression method involves adjusting the order of each step in the execution flow of the preceding subtask without changing the resource requirements set of the preceding subtask, reducing the execution simulation time required for the preceding subtask, and ensuring that the execution simulation end time of the compressed preceding subtask is earlier than the execution simulation start time of the subsequent subtask. The first sub-flow of the two subsequent subtasks executes within the time-series conflict interval, and the second sub-flow executes after the time-series conflict interval ends. The sum of the execution time windows of the two subsequent subtasks equals the original execution time window, and the resource requirements of the first sub-flow have no overlap with the resource requirements of the preceding subtask, while the resource requirements of the second sub-flow match the resource set released after the compression of the preceding subtask.

[0207] Continuing with the above example, the preceding subtask M (material inspection) is compressed: The original execution flow was "verify documents (10 minutes) → sample inspection (20 minutes) → record results (10 minutes)", with a total time of 40 minutes (14:00-14:40). It is adjusted to "verify documents and sample inspection in parallel (20 minutes) → record results (10 minutes)", with a compressed total time of 30 minutes (14:00-14:30). The end time of 14:30 is earlier than the start time of subtask N of 14:00 (original plan).

[0208] Subsequent subtask N (material warehousing) is split as follows: The original execution time window of 14:00-14:50 (50 minutes) is split into: the first sub-process (14:00-14:20): "Sorting the appearance of materials", with a resource requirement of 1 operator (not overlapping with the quality inspector of subtask M); the second sub-process (14:30-14:60): "Transporting and warehousing", with a resource requirement of 1 forklift (matching the forklift resources released after the compression of subtask M).

[0209] The total time for the two sub-processes is 50 minutes, consistent with the original time window.

[0210] Step 411: Perform time coordination between the execution flow of the preceding subtask and the execution flow of the following subtask to obtain the time coordination result.

[0211] Furthermore, the warehouse inventory control system coordinates the compressed process of the preceding subtask with the two sub-processes split from the subsequent subtask in terms of time, ensuring no time or resource conflicts. The coordination rules are as follows: the simulated end time of the compressed preceding subtask (e.g., 14:30) serves as the start time of the second sub-process of the subsequent subtask; the start time of the first sub-process of the subsequent subtask maintains a safe interval (i.e., does not overlap) with the start time of the compressed preceding subtask, and there are no resource conflicts.

[0212] Through collaboration, the timing of preceding and subsequent tasks can be smoothly connected, avoiding new conflicts.

[0213] Continuing with the above embodiment, the compressed execution time of the preceding subtask M is 14:00-14:30, and the end time is 14:30. The start time of the first subprocess of the subsequent subtask N, 14:00, is identical to the start time of subtask M, 14:00, but their resource requirements are for quality inspectors and operators respectively, which is conflict-free and conforms to the collaboration rules. The start time of the second subprocess of the subsequent subtask N is set to 14:30 (consistent with the end time of subtask M), achieving seamless connection. The time collaboration result is: M (14:00-14:30) → N1 (14:00-14:20) → N2 (14:30-14:60).

[0214] Step 412: Update the sequential dependency chain related to time planning conflicts in the execution simulation trigger sequence of the second target task based on the offset execution time window and time coordination results, so as to modify the trigger conditions of the tasks in the updated sequential dependency chain to trigger conditions based on the offset time window and process coordination relationship, ensuring that there are no time connection conflicts in the updated sequential dependency chain, and obtain the optimized execution strategy.

[0215] Furthermore, the warehouse inventory control system updates the sequential dependency chain based on the offset execution time window and time coordination results to form an optimized execution strategy. Specifically, it modifies the sequential dependency chain related to time planning conflicts in the simulated trigger sequence of the second target task execution, adjusting the task triggering conditions to be based on the offset time window and process coordination relationships (e.g., "the preceding task ends after compression → triggering the second sub-process of the subsequent task" or "the first sub-process of the subsequent task is triggered independently"). The updated sequential dependency chain must ensure no time connection conflicts, and the triggering logic of all tasks must conform to the optimized time plan.

[0216] Continuing with the above embodiment, the original sequential dependency chain is: M→N→O→P. It is updated to: M (14:00-14:30) trigger condition: time reaches 14:00; N1 (14:00-14:20) trigger condition: time reaches 14:00; N2 (14:30-14:60) trigger condition: M finishes execution (14:30); O (13:50-14:00, 14:30-14:50) trigger conditions: time reaches 13:50 and 14:30 respectively; P (14:30-14:50) trigger condition: 5 minutes after N2 starts (14:35). The updated sequential dependency chain has no time conflicts, forming an optimized execution strategy.

[0217] This invention, through compressing preceding tasks and splitting subsequent tasks, fully utilizes time-series conflict intervals, reducing total task time. Adjustments to the time offsets of affected subtasks prevent conflict propagation and ensure the rationality of overall time planning. The time coordination mechanism tightly connects the execution flows of preceding and subsequent tasks, matching resource utilization with time planning, eliminating previous time overlaps and improper connections. Therefore, it provides a systematic optimization path for tasks with time-series conflict types, ensuring smooth execution of material inbound and outbound tasks in VMI physical warehouses across time dimensions, and improving overall task efficiency and stability.

[0218] Furthermore, for tasks of the branch coordination conflict type, the process of optimizing the task based on the task nodes and task types of the conflicting tasks to obtain the optimized execution strategy includes:

[0219] Step 413: For the third objective task with branch coordination conflicts, based on the attribute differences of the association attributes of the conflicting branches corresponding to the third objective task, all affected subtasks within the scope of influence of the third objective task are classified into three task groups according to their respective branches and attribute characteristics.

[0220] Optionally, for the third objective task, the warehouse inventory control system determines the association attributes of its corresponding conflicting branches. These association attributes include resource requirements, execution time windows, branch information, node level, etc., and then identifies attribute differences. Further, based on the branch and attribute characteristics, the warehouse inventory control system divides all affected subtasks within its scope into three task groups: the first task group is the first source task group with attributes consistent with the first task in the third objective task; the second task group is the second source task group with attributes consistent with the second task in the third objective task; and the third task group is a cross-task group that possesses some attributes of both the first and second tasks.

[0221] In one embodiment, the third objective task consists of subtask A (belonging to branch 1, resource requirement is forklift F1, execution time window 8:00-9:00) and subtask B (belonging to branch 2, resource requirement is forklift F2, execution time window 8:30-9:30), which have a branch coordination conflict. The subtasks affected within its scope are: subtask C (belonging to branch 1, resource requirement is forklift F1, execution time window 7:30-8:30); subtask D (belonging to branch 2, resource requirement is forklift F2, execution time window 9:00-10:00); subtask E (belonging to branch 1, resource requirement is forklift F2, execution time window 8:20-9:20); and subtask F (belonging to branch 2, resource requirement is forklift F1, execution time window 8:40-9:40).

[0222] The main differences in the associated attributes of conflicting branches lie in resource requirements (F1 and F2) and execution time windows (8:00-9:00 and 8:30-9:30). The warehouse inventory control system is categorized as follows: First homogeneous task group (with attributes consistent with subtask A): Subtask C (resource requirement F1, belonging to branch 1); Second homogeneous task group (with attributes consistent with subtask B): Subtask D (resource requirement F2, belonging to branch 2); Cross-task group (possessing attributes from both A and B): Subtask E (belonging to branch 1, resource requirement F2), Subtask F (belonging to branch 2, resource requirement F1).

[0223] Step 414: Add intermediate nodes for branch coordination between conflicting branches. The resource requirements of the intermediate nodes for branch coordination are the differences in resource requirement types among attribute differences, the execution time window is the intersection of the time window feature differences among attribute differences, and the node level is the average level of the conflicting branches.

[0224] Furthermore, a branch coordination intermediate node is added between the conflicting branches to coordinate the task execution of the two branches. The attributes of the branch coordination intermediate node are as follows: resource requirement is the difference in resource requirement type among the differences in the associated attributes of the conflicting branches, that is, the different parts of the resource requirement type of the two branch tasks; execution time window is the intersection of the time window feature differences among the attribute differences, that is, the overlapping part of the execution time windows of the two branch tasks; node level is the average level of the conflicting branches, ensuring that the intermediate node is in an appropriate hierarchical position in the structure tree. Continuing the above embodiment, the resource requirement type difference of the conflicting branches is forklift F1 and forklift F2, the intersection of the time window feature differences is 8:30-9:00, the level of the conflicting branches is level two, and the average level is level two. A branch coordination intermediate node G is added: resource requirement is forklift F1 and forklift F2 (covering the difference in resource requirement type), execution time window is 8:30-9:00 (intersection of time windows), and node level is level two.

[0225] Step 415: Reassign each task in the third task group to a branch based on its attribute similarity with the conflicting branch to obtain the adjusted task group. Update the first and second source task groups based on the adjusted task groups to obtain the first and second adjusted task groups.

[0226] Furthermore, the warehouse inventory control system calculates the attribute similarity between each task in the third task group and the conflicting branches. Attribute similarity is determined based on the degree of matching of attributes such as resource requirements and execution time windows. Tasks in the third task group are then reassigned to their corresponding branches based on attribute similarity, resulting in an adjusted task group. The first and second source task groups are then updated based on the adjusted task groups, forming the first and second adjusted task groups.

[0227] Continuing with the above embodiment, subtask E (belonging to branch 1, resource requirement F2, execution time window 8:20-9:20) has a higher similarity in attributes to branch 2 (resource requirement F2); subtask F (belonging to branch 2, resource requirement F1, execution time window 8:40-9:40) has a higher similarity in attributes to branch 1 (resource requirement F1). The warehouse inventory control system reassigns subtask E to branch 2 and subtask F to branch 1. After adjustment: First adjusted task group: subtask C, subtask F; Second adjusted task group: subtask D, subtask E.

[0228] Step 416: Establish the first sequential dependency between the branch coordination intermediate node and each task in the first adjusted task group, and the second sequential dependency between the branch coordination intermediate node and each task in the second adjusted task group.

[0229] Furthermore, the warehouse inventory control system operates according to the sequential dependency rules: the execution simulation end time of all tasks in the first adjusted task group is used as the execution simulation start time of the branch coordination intermediate node, ensuring that the intermediate node starts execution after the first adjusted task group is completed; the execution simulation end time of the branch coordination intermediate node is used as the execution simulation start time of all tasks in the second adjusted task group, ensuring that the second adjusted task group starts execution after the intermediate node is completed, making the second adjusted task group a node connecting the two adjusted task groups.

[0230] Continuing with the above embodiment, in the first adjusted task group, the simulation end time for subtask C is 8:30, and the simulation end time for subtask F is 9:40. The latest end time, 9:40, is taken as the start time for the simulation of branch coordination intermediate node G. The execution time window for node G is 8:30-9:00, and the adjusted simulation end time is 9:00, which is taken as the start time for the simulation of the second adjusted task group. The established relationship is: First adjusted task group → Node G → Second adjusted task group. That is, after subtasks C and F are completed, node G begins execution; after node G is completed, subtasks D and E begin execution.

[0231] Step 417: Reconstruct the task execution sequence of conflicting branches based on the first and second order dependencies to obtain the optimized execution strategy.

[0232] Furthermore, the warehouse inventory control system reconstructs the task execution sequence of conflicting branches based on the established sequential dependency relationship. The optimized execution strategy follows the reconstruction rules: retain the original sequential dependency relationship within each conflicting branch, and replace the parallel conflict relationship between conflicting branches with the first or second sequential dependency relationship, that is, achieve the orderly execution of the two branch tasks by coordinating the intermediate nodes of the branches, thereby eliminating parallel conflicts.

[0233] Continuing with the above embodiment, in the original conflicting branch task execution sequence, subtask A of branch 1 and subtask B of branch 2 are in a parallel conflict relationship. After reconstruction, the execution sequence is: first adjusted task group (subtasks C, F) → branch coordination intermediate node G → second adjusted task group (subtasks D, E). The original sequential dependencies within branches 1 and 2 are retained, and the parallel conflict relationships between branches are replaced by sequential dependencies through node G. The resulting optimized execution strategy ensures that the tasks in both branches are executed in an orderly manner, without branch coordination conflicts.

[0234] The branch coordination intermediate node added in this embodiment of the invention acts as a good bridge, enabling two conflicting branch tasks to be connected in an orderly manner, avoiding parallel conflicts. The redistribution of tasks and the updating of task groups make task attribution more reasonable, reducing conflicts caused by mixed attributes. The reconstructed task execution sequence retains the original order within each branch, while standardizing the task execution order between branches through sequential dependencies, ensuring the smooth progress of the entire task flow, improving the execution efficiency and coordination of material inbound and outbound tasks in VMI physical warehousing, thereby enhancing the overall stability of warehousing operations.

[0235] Furthermore, the VMI warehouse inventory dynamic control system based on the Industrial Internet provided by the present invention will be described below. The VMI warehouse inventory dynamic control system based on the Industrial Internet described below can be referred to in correspondence with the VMI warehouse inventory dynamic control method based on the Industrial Internet described above.

[0236] Reference Figure 2 , Figure 2 This is a structural diagram of the VMI (Vendor Managed Inventory) dynamic control system for warehouses based on the Industrial Internet provided by this invention. The VMI dynamic control system for warehouses based on the Industrial Internet includes:

[0237] The twin model construction module 210 is used to construct a warehouse digital twin model that is proportionally mapped to the VMI physical warehouse based on the warehouse status information collected.

[0238] The structure tree construction module 220 is used to break down the received material inbound and outbound order information according to the order of task execution, obtain multiple sub-tasks, and construct an initial DOM task structure tree based on the multiple sub-tasks.

[0239] The task simulation module 230 is used to simulate the execution process of material inbound and outbound tasks in the warehouse digital twin model based on the initial DOM task structure tree, and to obtain conflicting tasks that are abnormal during the simulation process.

[0240] The strategy update module 240 is used to optimize tasks based on the task nodes and task types of conflicting tasks, obtain optimized execution strategies, and update the initial execution strategies corresponding to conflicting tasks in the initial DOM task structure tree based on the optimized execution strategies to obtain the target DOM task structure tree.

[0241] The warehouse inventory control module 250 is used to control the inbound and outbound of materials in the VMI physical warehouse based on the target DOM task structure tree.

[0242] This invention constructs a warehouse digital twin model that is proportionally mapped to the VMI physical warehouse, reflecting the real-time status of the VMI physical warehouse and avoiding decision-making errors caused by untimely understanding of the warehouse status. An initial DOM task structure tree is built based on multiple sub-tasks in the material inbound / outbound order information, clearly presenting the task execution logic and facilitating the detection of conflicting tasks. Then, the initial DOM task structure tree is used to perform pre-simulation in the warehouse digital twin model. Based on the task nodes and task types of conflicting tasks that appear abnormally during the simulation, task optimization is performed to obtain the target DOM task structure tree. Finally, warehouse inventory control is performed on the material inbound / outbound of the VMI physical warehouse based on the target DOM task structure tree. Therefore, through pre-simulation and optimization, conflicts in physical operations are avoided, ensuring the smooth execution of tasks in physical operations. This solves the problem of frequent material backlogs or stockouts in actual operations, reduces equipment waiting time and inefficient handling, improves material inbound / outbound efficiency, and thus improves overall warehouse operation efficiency.

[0243] Please see Figure 3 , Figure 3 An embodiment diagram of an electronic device provided in accordance with the present invention. For example... Figure 3 As shown, an embodiment of the present invention provides an electronic device 300, including a memory 310, a processor 320, and a computer program 311 stored in the memory 310 and executable on the processor 320. When the processor 320 executes the computer program 311, it implements the processes of steps 10 to 50.

[0244] Please see Figure 4 , Figure 4 An embodiment diagram of a computer-readable storage medium provided in accordance with an embodiment of the present invention is shown. Figure 4 As shown, this embodiment provides a computer-readable storage medium 400 on which a computer program 311 is stored. When the computer program 311 is executed by a processor, it implements the processes of steps 10 to 50.

[0245] On the other hand, the present invention also provides a computer program product, which includes a computer program that can be stored on a non-transitory computer-readable storage medium. When the computer program is executed by a processor, the computer can execute the VMI warehouse inventory dynamic control method based on the Industrial Internet provided by the above methods, which includes steps 10 to 50.

[0246] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, and not to limit them; although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features; and these modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A method for dynamic control of warehouse inventory based on the Industrial Internet, characterized in that, include: Based on the collected storage status information of the VMI physical warehouse, a storage digital twin model is constructed that is proportionally mapped to the VMI physical warehouse. The received material inbound / outbound order information is broken down into multiple subtasks according to the order of task execution. An initial DOM task structure tree is then constructed based on these subtasks. The attributes of each subtask include the type of preconditions, resource requirements, and execution time window. The task relationships between subtasks include sequential dependency, parallel compatibility, and branch triggering dimensions. Sequential dependency represents the mandatory relationship of tasks being executed in sequence, parallel compatibility represents the resource and time compatibility of tasks being executed simultaneously, and branch triggering represents the condition satisfaction of a task as the starting point of a branch. Specifically, it includes: Taking the total task in the material inbound / outbound order information as the root node, and the subtask with no prerequisite constraint type as the first subtask, perform sequential dependency dimension matching between the first subtask and the root node; if the dependency dimension matching result is consistent, determine whether there is a conflict between the resource requirements of the first subtask and the initial resource pool of the root node, and whether the execution time window of the first subtask includes the virtual start time of the root node; if there is no conflict in resource requirements and the execution time window includes the virtual start time, then the first subtask is determined as a first-level child node directly under the root node; For any two nodes in the first-level child nodes, if the two nodes satisfy resource compatibility and time compatibility, then the two nodes are determined to be parallel siblings, and all nodes with parallel sibling relationships are grouped into the same sibling group to obtain the sibling relationship group of the first-level child nodes. For each sibling relationship group, the lower-level child nodes are determined level by level based on the branch triggering dimension of the node. Candidate nodes of absolute prerequisite type and conditional prerequisite type are bound to the parent node according to the sequential dependency dimension of time order constraint and resource inheritance constraint to establish a sequential dependency chain within the same branch. Cross-validation of the parallel compatibility dimension is performed on the child nodes of the same level under different branches. Cross-branch parallel sibling relationships are established for nodes that meet the resource compatibility and time compatibility. This process continues until all child nodes are connected, resulting in the initial DOM task structure tree. Based on the initial DOM task structure tree, the execution process of material inbound and outbound tasks is simulated in the warehouse digital twin model to obtain conflicting tasks that exhibit anomalies during the simulation; specifically including: Establish an association mapping between the nodes corresponding to all subtasks in the initial DOM task structure tree and the material inbound / outbound operation units in the warehouse digital twin model to obtain the association mapping relationship; the association mapping rule is: the resource requirements of the nodes corresponding to the subtasks correspond to the resource units in the warehouse digital twin model, and the execution time window of the nodes corresponding to the subtasks correspond to the time axis in the warehouse digital twin model. Based on the aforementioned association mapping relationship and the sequential dependency chain under the same branch in the initial DOM task structure tree, the execution simulation of subtasks is triggered sequentially according to the sequential dependency relationship in the warehouse digital twin model to obtain the execution simulation trigger sequence; the triggering rule is: when the execution simulation status of the current subtask is completed, the execution simulation of the subsequent subtask is triggered. Based on the execution simulation trigger sequence, time connection conflicts of tasks in the sequential dependency chain are identified in the warehouse digital twin model to obtain time connection conflict task pairs; the identification rule is: if the execution simulation end time of the preceding subtask in the sequential dependency chain is later than the execution simulation start time of the following subtask, it is marked as a time connection conflict task pair. Based on the aforementioned association mapping relationship and the parallel sibling tasks belonging to the same sibling group in the initial DOM task structure tree, synchronous execution simulation monitoring is performed in the warehouse digital twin model to obtain conflicting parallel sibling task pairs; the monitoring content is: if the resource units corresponding to the parallel sibling tasks are occupied at the same time, and there is resource contention in the overlapping interval of the execution time windows of the parallel sibling tasks on the time axis, then they are marked as conflicting parallel sibling task pairs. Based on the aforementioned association mapping relationship and the execution simulation trigger sequence, combined with the cross-branch parallel tasks in the initial DOM task structure tree that have passed cross-branch parallel compatibility verification, resource occupation conflicts are detected in the warehouse digital twin model to obtain resource occupation conflict task pairs; the detection rule is: if a cross-branch parallel task is in the triggered execution state, and its corresponding resource unit is simultaneously requested and occupied during the execution simulation, it is marked as a resource occupation conflict task pair. The time-sequential conflict task pairs, conflicting parallel sibling task pairs, and resource-occupation conflict task pairs are identified as the conflicting tasks. Based on the task nodes and task types of the conflicting tasks, task optimization is performed to obtain an optimized execution strategy. Based on the optimized execution strategy, the initial execution strategy corresponding to the conflicting task is updated in the initial DOM task structure tree to obtain the target DOM task structure tree. Based on the target DOM task structure tree, the warehouse inventory is adjusted for the inbound and outbound of materials in the VMI entity warehouse.

2. The method for dynamic control of VMI warehouse inventory based on the Industrial Internet as described in claim 1, characterized in that, The dependency dimension matching result indicates that the virtual execution time of the root node is earlier than the start of the execution time window of the first subtask, and the virtual execution completion status of the root node is in the completed state; the resource compatibility indicates that the resource requirements of the two nodes do not overlap, and the time compatibility indicates that the execution time windows of the two nodes overlap.

3. The method for dynamic control of VMI warehouse inventory based on the Industrial Internet as described in claim 1, characterized in that, The steps to obtain the initial DOM task structure tree include: For each sibling relationship group of a first-level child node, nodes in the sibling relationship group with absolute or conditional precondition types are identified as candidate nodes. For a first candidate node whose prerequisite constraint type is absolute, if the node corresponding to the prerequisite task pointed to by the subtask of the first candidate node is a first-level sub-node, then the first candidate node is determined as a second-level sub-node. For a second candidate node with a precondition type of precondition, if the execution result of the subtask corresponding to the first-level child node satisfies the triggering condition of the subtask corresponding to the second candidate node, then the second candidate node is determined as a second-level child node. For each second-level child node, the second-level child node is bound to its corresponding first-level child node according to the sequential dependency dimension based on the time order constraint and the resource inheritance constraint, and the sequential dependency binding result is obtained. The time order constraint indicates that the end of the execution time window of the parent node is earlier than the start of the execution time window of the child node, and the resource inheritance constraint indicates that the resource release set of the parent node contains the initial resource requirements of the child node. The initial DOM task structure tree is constructed based on the sequential dependency binding results of the second-level child nodes.

4. The method for dynamic control of VMI warehouse inventory based on the Industrial Internet as described in claim 3, characterized in that, The initial DOM task structure tree is constructed based on the sequential dependency binding results of second-level child nodes, including: For all second-level child nodes under the same parent node, if the sequential dependency binding result between two second-level child nodes satisfies the sequential dependency binding condition, then construct the sequential dependency chain between the two second-level child nodes within the same branch according to the task order of the two second-level child nodes. For all child nodes under different parent nodes, perform cross-validation on two child nodes at the same level under different branches to obtain the cross-validation results on the parallel compatibility dimension. The cross-validation on the parallel compatibility dimension includes whether the resource requirements of two child nodes at the same level under different branches have no overlap, and whether the execution time windows of the two child nodes overlap. If the parallel compatibility dimension verification result satisfies both resource compatibility and time compatibility, then establish a cross-branch parallel sibling relationship between two child nodes at the same level under different branches; Based on the sequential dependency chain of two second-level child nodes within the same branch in each level, the sibling relationship group in each level, and the cross-branch parallel sibling relationship between two child nodes of the same level under different branches in each level, establish the mutual connection relationship between all child nodes until all child nodes have corresponding connections, and obtain the initial DOM task structure tree.

5. The method for dynamic control of VMI warehouse inventory based on the Industrial Internet according to any one of claims 1 to 4, characterized in that, The specific process for determining the task type of the conflicting task includes: The scope of influence of the conflicting tasks is defined based on their associated attributes. These attributes include the intersection of resource requirements, the intersection of execution time windows, branch information, and node level in the initial DOM task structure tree. The definition rules are as follows: if the conflicting tasks belong to the same sequential dependency chain, the scope of influence includes all subsequent subtasks in the sequential dependency chain, as well as associated subtasks related to the sequential dependency chain in the execution simulation trigger sequence; if the conflicting tasks belong to parallel tasks in different branches, the scope of influence includes all subtasks in the two branches that share resources or overlap in time with the conflicting tasks, as well as associated subtasks related to the two branches in the execution simulation trigger sequence. If the intersection of the resource requirements of the conflicting tasks is non-empty, and the resource requirements of all affected subtasks within the scope of influence are included in the intersection of resource requirements, then the task type of the conflicting task is determined to be a resource allocation conflict type. If the intersection of the execution time windows of the conflicting tasks is non-empty, and the execution time windows of all affected subtasks within the scope of influence overlap with the intersection of the execution time windows, and the execution simulation end time of the preceding subtask is later than the execution simulation start time of the following subtask, then the task type of the conflicting task is determined to be a time planning conflict type. If the branch information of the conflicting task does not match, and all affected subtasks within the scope of influence belong to two different branches, and the cross-branch parallel compatibility result does not meet resource compatibility and / or time compatibility, then the task type of the conflicting task is determined to be a branch coordination conflict type.

6. A VMI (Vendor Managed Inventory) dynamic control system based on the Industrial Internet, characterized in that, include: The twin model construction module, the structure tree construction module, the task simulation module, the strategy update module, and the warehouse inventory control module are used to implement the VMI warehouse inventory dynamic control method based on the Industrial Internet as described in any one of claims 1 to 5.

7. An electronic device, comprising: Memory, used to store computer software programs; A processor for reading and executing the computer software program, characterized in that, when the processor executes the computer software program, it implements the VMI warehouse inventory dynamic control method based on the Industrial Internet as described in any one of claims 1 to 5.

8. A non-transitory computer-readable storage medium, wherein a computer software program is stored therein, characterized in that, When the computer software program is executed by the processor, it implements the VMI warehouse inventory dynamic control method based on the Industrial Internet as described in any one of claims 1 to 5.

Images