Intelligent shared pallet automated warehouse-in and warehouse-out method

Abstract

The present application relates to the technical field of intelligent logistics, and in particular to an automatic warehouse-in and warehouse-out method for intelligent shared pallets, which comprises collecting pallet operation state data through a multi-source sensing module and preprocessing, generating task decomposition parameters based on a task decomposition model, constructing a multi-objective planning model for global scheduling, and outputting execution instructions through a hierarchical execution control model. The present application can realize efficient and accurate warehouse-in and warehouse-out management of shared pallets, optimize the flow efficiency and resource conflicts, and improve the intelligent level and operation efficiency of the logistics system.

Description

TECHNICAL FIELD

[0001] The present application belongs to the technical field of logistics and warehouse automation, specifically an intelligent shared pallet automated warehouse-in and warehouse-out method. BACKGROUND

[0002] Under the background of continuous improvement of automation level in today's logistics and warehouse industry, intelligent and efficient warehouse-in and warehouse-out management methods have gradually become the core direction of research. As an innovative logistics carrier, intelligent shared pallets show significant advantages in improving warehouse efficiency and reducing labor costs. However, the existing automated warehouse-in and warehouse-out management systems and technical solutions still have certain limitations in the intelligent management and scheduling of shared pallets, and it is difficult to meet the demand for efficient flow of shared pallets in modern logistics.

[0003] A kind of pipe material class automatic stereoscopic warehouse and warehouse-in and warehouse-out method with publication number CN110626700B, which realizes the automatic storage and warehouse-in and warehouse-out operation of long heavy steel pipe material through the collaborative work of material frame conveying trolley, lifting mechanism, walking mechanism and storage and retrieval mechanism. However, this technical solution is mainly designed for specific types of goods (such as pipe material), and lacks general support for shared pallets. In addition, this solution does not involve intelligent scheduling and dynamic allocation strategy of pallets, which may affect the flow efficiency of pallets in the warehouse process, and it is difficult to adapt to the high-frequency and multi-variety warehouse-in and warehouse-out demand in modern logistics.

[0004] A kind of mobile multi-unmanned aerial vehicle automatic intelligent warehouse-in and warehouse-out system and warehouse-in and warehouse-out method with publication number CN112707072B, which realizes the automatic storage and scheduling of multi-unmanned aerial vehicles through the cooperation of lifting shaft frame, stereoscopic hangar frame and warehouse-in and warehouse-out manipulator system. However, this technical solution is mainly applied to the storage scene of unmanned aerial vehicles, and does not fully consider the special needs of shared pallets in the warehouse environment, such as dynamic sharing of pallets, state monitoring and seamless integration with logistics systems. In addition, this solution has weak automatic identification and positioning ability for pallets, which may have certain impact on the accuracy and efficiency of pallet warehouse-in and warehouse-out.

[0005] The above problems show that the existing automated warehouse-in and warehouse-out management systems and technical solutions have certain limitations in the application scene of intelligent shared pallets, especially in the intelligent scheduling, dynamic allocation of pallets and integration with logistics systems. Therefore, the present application provides an intelligent shared pallet automated warehouse-in and warehouse-out method, which aims to optimize the warehouse-in and warehouse-out process of pallets, improve the flow efficiency of pallets, and realize efficient cooperation with logistics systems, so as to better meet the demand for efficient management of intelligent shared pallets in modern logistics. SUMMARY

[0006] The present application aims to provide an intelligent shared pallet automated warehouse entry and exit method to solve the problems raised in the background art.

[0007] To achieve the above-mentioned purpose, the present application provides the following technical solutions: an intelligent shared pallet automated warehouse entry and exit method, the method comprising:

[0008] The running state data of the shared pallet is collected by a multi-source perception module, which includes a pallet identification recognition unit, a pallet position tracking unit and a pallet load monitoring unit; the running state data of the shared pallet is preprocessed based on a hierarchical analysis framework to obtain a standardized state set; the standardized state set is input into a pre-configured task deconstruction model, which adopts a partition scheduling mechanism and performs hierarchical division on the warehouse entry and exit tasks based on business logic rules to generate task deconstruction parameters;

[0009] A multi-objective planning model is constructed according to the task deconstruction parameters, which takes the highest flow efficiency and the least resource conflict as the optimization target, and uses a dynamic allocation algorithm to perform global scheduling on the warehouse entry and exit tasks, wherein the dynamic allocation algorithm introduces task priority identification and resource adaptation degree evaluation; an optimal scheduling scheme is output based on the multi-objective planning model;

[0010] A hierarchical execution control model is established according to the optimal scheduling scheme, which includes a strategy layer, a coordination layer and an execution layer, wherein the strategy layer performs global task planning based on the task deconstruction parameters, the coordination layer performs local resource adjustment based on the optimal scheduling scheme, and the execution layer realizes task tracking of the logistics nodes based on a load balancing algorithm; execution control instructions are output through the hierarchical execution control model to realize efficient warehouse entry and exit management of the shared pallet.

[0011] Preferably, the standardized state set is input into a pre-configured task deconstruction model, which adopts a partition scheduling mechanism and performs hierarchical division on the warehouse entry and exit tasks based on business logic rules to generate task deconstruction parameters, including:

[0012] The standardized state set is obtained, which includes pallet identification information, pallet position information, pallet load state and cross-system association identifier; a state feature space is constructed based on the standardized state set, and a resource capability space is constructed based on the transportable resource amount, storage capacity and network communication bandwidth of the logistics platform;

[0013] construct a multi-dimensional business logic rule based on the state feature space and the resource capability space, the multi-dimensional business logic rule including a task type matching item, a processing timeliness requirement item, a resource adaptation item, and a cross-system collaboration item, wherein the task type matching item is obtained by evaluating the compatibility of a task attribute and a processing module, the processing timeliness requirement item is calculated by matching the task update frequency and the completion time, the resource adaptation item is evaluated by the task size and the resource load balancing degree, and the cross-system collaboration item is determined by the correspondence relationship between the association identifier and the collaborative processing flow;

[0014] construct a partition scheduling mechanism, the partition scheduling mechanism including an initial task set, a hierarchical division rule, and an association verification module, the initial task set containing a plurality of warehouse-in and warehouse-out task units, the hierarchical division rule performing hierarchical sorting on the task units according to the business process priority, and the association verification module verifying the dependency relationship between the task units based on the multi-dimensional business logic rule;

[0015] update the initial task set by using an iterative optimization method, retain high-priority task units by priority screening, merge low-priority task units and split complex task units, calculate the hierarchical distribution and resource demand of the warehouse-in and warehouse-out tasks based on the updated task set, and output task deconstruction parameters based on the partition scheduling mechanism, the task deconstruction parameters including the task hierarchical distribution, the resource demand list, the task unit dependency relationship, and the cross-system collaboration identifier.

[0016] Preferably, a multi-objective planning model is constructed according to the task deconstruction parameters, the multi-objective planning model taking the highest flow transfer efficiency and the least resource conflict as optimization objectives, a dynamic allocation algorithm is used to perform global scheduling on the warehouse-in and warehouse-out tasks, and an optimal scheduling scheme is output based on the multi-objective planning model, including:

[0017] construct a planned multi-objective function, the multi-objective function including a flow transfer efficiency optimization objective function and a resource conflict optimization objective function, wherein the flow transfer efficiency optimization objective function is calculated by summing the ratio of the task unit processing time to the corresponding resource processing speed, and the resource conflict optimization objective function is calculated by weighted summing the transportation resource occupation rate, the storage resource occupation rate, and the network communication occupation rate of each logistics node;

[0018] construct a planning constraint condition based on the multi-objective function, the planning constraint condition including a time constraint, a resource constraint, a priority constraint, and a collaboration constraint, the time constraint being used to limit the completion time range of the warehouse-in and warehouse-out tasks, the resource constraint being used to limit the upper limit of the resource usage of each logistics node, the priority constraint being used to limit the scheduling order of the high-priority task units, and the collaboration constraint being used to ensure the processing order consistency of the cross-system task units;

[0019] The node sequence is used to encode the warehouse in and out task, each node contains task unit identification and corresponding logistics node information, the task transfer cost and heuristic evaluation value are used to construct the task allocation probability, and the next feasible node is determined through the task allocation probability; wherein the heuristic evaluation value is obtained by weighted calculation of the processing time difference and resource occupation difference between nodes;

[0020] An adaptive task transfer cost mechanism is introduced, which uses a time-varying weight coefficient that decreases linearly with the increase of the number of scheduling iterations, and dynamically adjusts the task transfer cost through the time-varying weight coefficient;

[0021] A dynamic heuristic evaluation mechanism is introduced, which uses a time-varying influence factor that increases linearly with the increase of the number of scheduling iterations, and adjusts the weight of the heuristic evaluation value in the task allocation probability through the time-varying influence factor; based on the adaptive task transfer cost mechanism and the dynamic heuristic evaluation mechanism, iterative optimization is carried out, the scheduling scheme generated by each iteration is evaluated in terms of flow transfer cost and resource cost, and non-dominated solutions are added to the Pareto front solution set;

[0022] The optimal solution that meets the flow transfer-resource trade-off is selected from the Pareto front solution set as the optimal scheduling scheme, and the optimal scheduling scheme is processed to generate the optimal node sequence and the corresponding logistics node resource allocation parameters.

[0023] Preferably, the strategy layer performs global task planning based on the task decomposition parameters, which includes:

[0024] The task processing path is described by parameterized flow, and the task processing path is represented as a function of flow parameters, the value range of the flow parameters is 0 to 1, and the task processing path includes task input stage, intermediate processing stage and result output stage;

[0025] The task processing path is described based on piecewise linear function, and the time schedule of the task processing path is calculated by summing the product of node coordinates and piecewise weights, and the piecewise weights are calculated by the ratio of task unit priority and processing complexity;

[0026] The task processing path constraint condition is constructed, which includes time constraint, sequence constraint, resource constraint and quality constraint, the time constraint is used to limit the time schedule range of the task processing path, the sequence constraint is used to limit the execution order of the task unit, the resource constraint is used to limit the upper limit of resource use of each stage, and the quality constraint is used to limit the accuracy requirement of task processing result;

[0027] A global multi-objective optimization function is constructed, which includes a total process duration item, a resource usage fluctuation item, a sequence deviation item, and a quality compliance item, and each item in the global multi-objective optimization function is weighted and combined by a weighting coefficient;

[0028] The task processing path interval is discretized into multiple process segments, the global multi-objective optimization function is discretized, a global discretized objective function is constructed, and the global discretized objective function includes a process segment duration, a resource usage, a sequence deviation value, and a quality compliance rate;

[0029] The global discretized objective function is iteratively optimized by using a gradient descent method, the gradient value of the objective function with respect to the control node is calculated, and the time schedule of the control node is updated along the negative gradient direction;

[0030] The optimized task processing path is linearly interpolated and smoothed, a smooth and continuous task processing path is generated by maintaining the continuity of the time schedule, resource usage, and sequence relationship at the interpolation endpoints, and a node sequence and corresponding logistics node processing speed are generated based on the smooth and continuous task processing path.

[0031] Preferably, the coordination layer performs local resource adjustment based on the optimal scheduling scheme, which includes:

[0032] A local adjustment window is constructed based on the current load, current resource occupation, and current processing progress of the logistics platform, the size of the local adjustment window is adaptively adjusted by a load correlation coefficient, and the size of the local adjustment window is positively correlated with the size of the current load;

[0033] A local resource model is constructed using real-time monitoring data, standardized resource occupation data is obtained by normalizing resource usage data, the availability identifier of the resource state table is updated based on the standardized resource occupation data, and the resource availability probability value of each logistics node is calculated by a probability accumulation method;

[0034] An exponential smoothing algorithm is used to predict dynamic resource demand, a future resource demand state vector is predicted based on historical demand data, the predicted state is updated based on real-time monitoring data, and an accurate prediction and change trend of resource demand are obtained;

[0035] A resource adjustment model is constructed, the resource allocation rules of the logistics platform are taken as a state equation, the state equation includes transportation resources, storage resources, and network communication resources, a state constraint condition and an allocation constraint condition are constructed, the state constraint condition is used to limit the value range of resource occupation and the value range of processing progress, and the allocation constraint condition is used to limit the value range of resource allocation adjustment amplitude and adjustment frequency;

[0036] A multi-objective cost function is constructed, which includes a reference schedule tracking term, a resource conflict avoidance term, a process smoothing term, and an energy consumption term, and each term in the multi-objective cost function is combined by a weighting coefficient;

[0037] The multi-objective cost function is optimized and solved by using a Lagrange multiplier method, a Lagrange function is constructed and a constraint condition is introduced, and an optimal adjustment amount is obtained by solving a partial derivative equation set.

[0038] Preferably, the execution layer implements task tracking of the logistics node based on a load balancing algorithm, and outputs an execution control instruction through the hierarchical execution control model to realize efficient in-out warehouse management of the shared pallet, which includes:

[0039] A multi-node resource allocation model of the logistics platform is established, which includes a transportation resource allocation equation and a storage resource allocation equation, the transportation resource allocation equation includes a task processing amount, a node transportation capacity, and a task priority, and the storage resource allocation equation includes a data storage amount, a node storage capacity, and a data access frequency;

[0040] The multi-node resource allocation model is constructed as a state space expression, a state vector of the state space expression includes a transportation resource occupation rate, a storage resource occupation rate, a task processing progress, and a node load balancing degree, and a control vector of the state space expression includes a transportation resource adjustment amount and a storage resource adjustment amount;

[0041] The state space expression is linearized, the partial derivatives of the system state equation with respect to the state vector and the control vector are calculated, and a linearized prediction model is constructed;

[0042] A task tracking prediction cost function is constructed, which includes a tracking error term, a control amount penalty term, and a control increment penalty term, and each penalty term is combined by a weight matrix;

[0043] A state constraint condition is constructed, which includes a resource occupation rate constraint and a task processing progress constraint, a control constraint condition is constructed, which includes a transportation resource adjustment amount constraint and a storage resource adjustment amount constraint, and a control increment constraint condition is constructed, which includes a transportation resource increment constraint and a storage resource increment constraint;

[0044] The task tracking prediction cost function is converted into a standard form of a quadratic programming problem, a quadratic matrix and a linear term coefficient are calculated, and an inequality constraint matrix and an equality constraint matrix are constructed;

[0045] The quadratic programming problem is solved by using an efficient set method, the optimal solution is gradually approached by identifying effective constraints and solving sub-problems.

[0046] The control quantity mapping is performed based on the optimization solution result, the total transportation resource adjustment amount is distributed to each logistics node through a transportation resource distribution matrix, and the total storage resource adjustment amount is distributed to each node storage unit through a storage resource distribution matrix;

[0047] The resource distribution of the logistics node is subjected to amplitude limiting processing, the transportation resource distribution amount is subjected to amplitude limiting according to the maximum transportation capacity of the node, the storage resource distribution amount is subjected to amplitude limiting according to the maximum storage capacity of the node, and a final execution control instruction is generated.

[0048] Preferably, the running state data of the shared pallet is preprocessed based on the hierarchical analysis framework to obtain a standardized state set, including: performing format conversion processing on the shared pallet running state data collected by the pallet identification recognition unit, the pallet position tracking unit and the pallet load monitoring unit respectively to obtain a unified data format; performing outlier cleaning on each format data, calculating the mean and variance of the data through a sliding window, and eliminating abnormal data outside the range; extracting the core business information of each data using a keyword matching algorithm, and establishing a corresponding relationship between different system data through information association; converting the associated business information to a unified data model, fusing multiple system data information through a weighted average method, and generating a standardized state set containing a time stamp.

[0049] Preferably, the partition scheduling mechanism is constructed, including: initializing a task set using uniform distribution, each task unit containing the business type and time requirement of the warehouse-in and warehouse-out task; calculating the node resource demand corresponding to the task unit through the logistics platform resource distribution rule, mapping the task unit demand to the resource capacity table based on the interface parameter configuration, and generating a task unit resource demand list; calculating the matching error of the task unit resource demand and the actual resources of the node, and constructing a business logic rule based on the error sum of squares; obtaining the priority value of each task unit through normalization processing, retaining the task unit with a priority value greater than a preset threshold, and eliminating the task unit with a priority value less than the preset threshold.

[0050] Preferably, the task transfer cost is constructed, including: calculating the processing time difference between the current node and the next node as the processing time cost; calculating the adjustment change amount of each resource required by the logistics platform from the current node to the next node as the resource adjustment cost; calculating the minimum distance of resource conflict in the scheduling scheme, and increasing the penalty cost if it is less than a safety threshold; processing the processing time cost, the resource adjustment cost and the penalty cost through linear combination to generate the total task transfer cost.

[0051] Preferably, solving the quadratic programming problem by using the active set method comprises: initializing an active constraint set, adding all constraints to a candidate set; solving an initial sub-problem to obtain an initial feasible solution; calculating constraint violation of the current solution, adding the constraint with the largest violation to the active set; re-solving the sub-problem to update the current solution; repeating the above steps until the violation of all constraints is less than a preset precision to obtain an optimal solution.

[0052] The application provides an intelligent shared pallet automated warehouse-in and warehouse-out method, aiming to optimize the pallet warehouse-in and warehouse-out process, improve the pallet circulation efficiency, and realize efficient cooperation with the logistics system, so as to better meet the demand of modern logistics for efficient management of intelligent shared pallets. BRIEF DESCRIPTION OF DRAWINGS

[0053] Figure 1 The figure is a schematic diagram of the overall process of the intelligent shared pallet automated warehouse-in and warehouse-out method in the embodiment of the application.

[0054] Figure 2 The figure is a schematic diagram of the partition scheduling mechanism of the task decomposition model in the embodiment of the application.

[0055] Figure 3 The figure is a flowchart of the dynamic allocation algorithm of the multi-objective programming model in the embodiment of the application.

[0056] Figure 4 The figure is a schematic diagram of the three-layer structure of the hierarchical execution control model in the embodiment of the application.

[0057] Figure 5 The figure is a schematic diagram of the linearization processing and quadratic programming solving process of the state space expression in the embodiment of the application. DETAILED DESCRIPTION

[0058] The application provides an intelligent shared pallet automated warehouse-in and warehouse-out method, combined with the attached Figure 1 to the attached Figure 5 The specific embodiments of the application are described in detail. The attached Figure 1 shows the complete process from data collection of the multi-source perception module to output of the execution control instruction of the hierarchical execution control model, the attached Figure 2 to the attached Figure 5 respectively show the specific implementation logic of the task decomposition model, the multi-objective programming model, the hierarchical execution control model and the state space expression.

[0059] In the specific implementation process, first, a standardized state set is obtained through data collection and preprocessing operations of the multi-source perception module. The multi-source perception module includes a pallet identification recognition unit, a pallet position tracking unit and a pallet load monitoring unit, which are respectively responsible for collecting identification information, position information and load state information of the pallet.

[0060] The pallet identification unit uses RFID technology to read the unique code on the pallet and convert it into a digital signal; the pallet location tracking unit uses ultra-wideband positioning technology to obtain the three-dimensional coordinates of the pallet in the warehouse in real time; and the pallet load monitoring unit uses pressure sensors to detect the weight distribution of goods on the pallet. The data acquisition frequency of the above three units is dynamically adjusted according to the operational needs of the logistics platform to ensure the real-time performance and accuracy of the data.

[0061] The collected data is converted into a standard data format, then outliers are removed using a sliding window algorithm, and core business information is extracted using a keyword matching algorithm. For example, the unique code in the pallet identification information is linked to the coordinate data in the pallet location information through a timestamp, thus forming a standardized status set that includes pallet identification, location, and load status.

[0062] After data acquisition and preprocessing, the standardized state set is input into the task deconstruction model for hierarchical partitioning and task deconstruction. (See attached image) Figure 2 As shown, the task deconstruction model adopts a partitioned scheduling mechanism. Its logical structure includes an initial task set, hierarchical partitioning rules, and an association verification module. The initial task set consists of multiple inbound and outbound task units. Each task unit contains attributes such as tray identification information, task type, processing time requirements, and resource requirement list.

[0063] The hierarchical division rules prioritize task units based on business process priorities. For example, higher-priority task units are allocated resources and their processing order is arranged first. The association verification module is used to verify the dependencies between task units. For example, an inbound task for a certain pallet can only be executed after its preceding outbound task has been completed.

[0064] To optimize the task deconstruction process, multi-dimensional business logic rules were constructed, including task type matching items, processing time requirements items, resource adaptation items, and cross-system collaboration items. These rules are implemented through the construction of a state feature space and a resource capability space. The state feature space describes the pallet's operating state, while the resource capability space reflects the logistics platform's transportation resources, storage capacity, and network communication bandwidth.

[0065] The initial task set is updated through an iterative optimization method, retaining high-priority task units and merging low-priority task units. The final generated task deconstruction parameters include task hierarchy distribution, resource requirement list, task unit dependency relationship, and cross-system collaboration identifier.

[0066] After task decomposition is completed, the global scheduling phase of the multi-objective programming model begins. (See attached image) Figure 3As shown, the core of the multi-objective programming model is the optimization target of the highest flow efficiency and the least resource conflict. In order to achieve this goal, first, a multi-objective function is constructed, including a flow efficiency optimization objective function and a resource conflict optimization objective function.

[0067] The flow efficiency optimization objective function is calculated by summing up the ratio of the task unit processing time to the corresponding resource processing speed, and the resource conflict optimization objective function is calculated by weighted summing up the transportation resource occupation rate, the storage resource occupation rate and the network communication occupation rate of each logistics node. On this basis, a dynamic allocation algorithm is introduced to globally schedule the in-and-out warehouse tasks.

[0068] The key of the dynamic allocation algorithm lies in the calculation of the task allocation probability. A heuristic evaluation value is obtained by weighted calculation of the processing time difference and the resource occupation difference between nodes, and then the next feasible node is determined. At the same time, an adaptive task transfer cost mechanism and a dynamic heuristic evaluation mechanism are introduced to improve the scheduling efficiency. The adaptive task transfer cost mechanism adopts a time-varying weight coefficient which linearly decays with the increase of the number of scheduling iterations, and the dynamic heuristic evaluation mechanism adopts a time-varying influence factor which linearly grows with the increase of the number of scheduling iterations.

[0069] Through the synergistic effect of the two mechanisms, the scheduling scheme generated by each iteration is evaluated in terms of flow cost and resource cost, and the non-dominated solution is added to the Pareto frontier solution set. Finally, the optimal solution that meets the flow-resource trade-off is selected from the Pareto frontier solution set as the optimal scheduling scheme, and the optimal node sequence and the corresponding logistics node resource allocation parameters are generated.

[0070] The hierarchical execution control model is the core execution part of the present application, as shown in FIG. 2. Figure 4 As shown, the three-layer structure includes a strategy layer, a coordination layer and an execution layer. The strategy layer performs global task planning based on task decomposition parameters, adopts parameterized process to describe the task processing path, and represents the task processing path as a function of process parameters. A piecewise linear function is used to describe the time schedule of the task processing path, and the specific time schedule value is calculated by summing up the product of the node coordinates and the piecewise weights.

[0071] The piecewise weight is determined by the ratio of the task unit priority to the processing complexity, so as to ensure that high-priority task units obtain more resource allocation. The coordination layer performs local resource adjustment based on the optimal scheduling scheme, and realizes dynamic allocation of resources by constructing a local adjustment window and a local resource model. The size of the local adjustment window is determined by the current load of the logistics platform, and the higher the load, the larger the window size. The local resource model normalizes the resource usage data and updates the availability identifier of the resource state table, and simultaneously uses an exponential smoothing algorithm to predict the state vector of future resource demand.

[0072] The resource adjustment model is solved by Lagrange multiplier method, a Lagrange function is constructed and constraint conditions are introduced, and finally the optimal adjustment amount is obtained. The execution layer realizes task tracking of the logistics node based on a load balancing algorithm, and allocates transportation resources and storage resources through the multi-node resource allocation model. The transportation resource allocation equation contains the task processing amount, the node transportation capacity and the task priority, and the storage resource allocation equation contains the data storage amount, the node storage capacity and the data access frequency. The state space expression is used to describe the resource allocation process of the execution layer, and the state vector includes the transportation resource occupation rate, the storage resource occupation rate, the task processing progress and the node load balancing degree, and the control vector includes the transportation resource adjustment amount and the storage resource adjustment amount.

[0073] As shown in the accompanying drawings, Figure 5 After linearization, the state space expression constructs a linear prediction model, and the task tracking prediction cost function includes a tracking error term, a control amount penalty term and a control increment penalty term. The quadratic programming problem is solved by the efficient set method, and the optimal solution is gradually approached to generate the final execution control instruction.

[0074] In practical applications, the method of the present application can be deployed in a large logistics warehouse center, real-time collection of pallet running state data is performed by a multi-source perception module and pretreatment is performed, then task decomposition parameters are generated by a task decomposition model, an optimal scheduling scheme is output by a multi-objective planning model, and finally efficient warehouse-in and warehouse-out management is realized by a hierarchical execution control model.

[0075] For example, in a certain scenario, when a batch of pallets need to complete warehouse-in and warehouse-out operations at the same time, the multi-source perception module first collects the identification, position and load state information of all pallets and generates a standardized state set. The task decomposition model generates a task level distribution according to the standardized state set, arranges high-priority warehouse-out tasks in front, and low-priority warehouse-in tasks at the back.

[0076] The multi-objective planning model calculates the optimal scheduling scheme by a dynamic allocation algorithm to ensure reasonable allocation of transportation resources and storage resources. The hierarchical execution control model generates execution control instructions according to the optimal scheduling scheme to guide the logistics node to complete specific warehouse-in and warehouse-out operations. In the whole process, seamless connection is realized between modules through data interfaces to ensure the accuracy and timeliness of information transmission.

[0077] In order to better enable relevant persons in the art to fully understand and implement the present application, the specific implementation principles of the present application are further supplemented below in conjunction with a specific application scenario.

[0078] In a practical application at a logistics warehousing center, when a batch of shared pallets needs to complete both inbound and outbound operations simultaneously, the pallet identification unit in the multi-source sensing module first reads the unique code on each pallet using radio frequency identification (RFID) technology and converts it into a digital signal. The pallet location tracking unit acquires the three-dimensional coordinate information of the pallets in real time based on ultra-wideband positioning technology, while the pallet load monitoring unit uses pressure sensors to detect the weight distribution of goods on the pallets.

[0079] The data collection frequency is dynamically adjusted according to the current operational needs of the logistics platform to ensure the real-time nature and accuracy of the data. Subsequently, the collected data undergoes format conversion processing to be unified into a standard data format, and outliers are removed using a sliding window algorithm, ultimately forming a standardized status set that includes pallet identification, location, and load status.

[0080] After generating a standardized set of states, the task deconstruction model hierarchically divides the set and deconstructs the tasks. The initial task set consists of multiple inbound and outbound task units, each of which contains attributes such as pallet identification information, task type, processing time requirements, and resource requirement list.

[0081] The partitioned scheduling mechanism prioritizes task units based on business process priorities. For example, high-priority outbound tasks are executed first, while low-priority inbound tasks are merged or delayed. The association verification module verifies the dependencies between task units; for example, an inbound task for a certain pallet can only be executed after its preceding outbound task has been completed.

[0082] The initial task set is updated through an iterative optimization method, retaining high-priority task units and merging low-priority task units. The final generated task deconstruction parameters include task hierarchy distribution, resource requirement list, task unit dependency relationship, and cross-system collaboration identifier.

[0083] After task decomposition, the process enters the global scheduling phase of the multi-objective programming model. This model optimizes for both maximizing workflow efficiency and minimizing resource conflicts, constructing both workflow efficiency and resource conflict optimization objective functions.

[0084] The objective function for optimizing the flow efficiency is calculated by summing the ratios of the processing time of the task unit to the processing speed of the corresponding resource, while the objective function for optimizing resource conflicts is calculated by weighted summing of the transportation resource occupancy rate, storage resource occupancy rate, and network communication occupancy rate of each logistics node.

[0085] On this basis, the dynamic allocation algorithm obtains a heuristic evaluation value through the weighted calculation of the processing time difference and the resource occupation difference between nodes, and then determines the next feasible node. The adaptive task transfer cost mechanism and the dynamic heuristic evaluation mechanism work together, and the scheduling scheme generated by each iteration is evaluated in terms of flow transfer cost and resource cost. Non-dominated solutions are added to the Pareto frontier solution set. Finally, the optimal solution that meets the flow transfer-resource trade-off is selected from the Pareto frontier solution set as the optimal scheduling scheme, and the optimal node sequence and the corresponding logistics node resource allocation parameters are generated.

[0086] The hierarchical execution control model is the core execution part of the application, and its three-layer structure includes a strategy layer, a coordination layer and an execution layer. The strategy layer performs global task planning based on task decomposition parameters, uses a segmented linear function to describe the time schedule of the task processing path, and calculates the specific time schedule value by summing the product of the node coordinates and the segmented weights.

[0087] The segmented weight is determined by the ratio of the task unit priority to the processing complexity, so as to ensure that high-priority task units obtain more resource allocation. The coordination layer performs local resource adjustment based on the optimal scheduling scheme, and realizes dynamic allocation of resources by constructing a local adjustment window and a local resource model.

[0088] The size of the local adjustment window is determined by the current load of the logistics platform, and the higher the load, the larger the window size. The local resource model normalizes the resource usage data and updates the availability identifier of the resource state table, and simultaneously uses an exponential smoothing algorithm to predict the state vector of future resource demand. The resource adjustment model is optimized and solved by the Lagrange multiplier method, a Lagrange function is constructed and a constraint condition is introduced, and finally the optimal adjustment amount is obtained.

[0089] The execution layer realizes task tracking of logistics nodes based on a load balancing algorithm, and allocates transportation resources and storage resources through a multi-node resource allocation model. The transportation resource allocation equation includes the processing task amount, the node transportation capacity and the task priority, and the storage resource allocation equation includes the data storage amount, the node storage capacity and the data access frequency.

[0090] In the execution layer, the state space expression is used to describe the resource allocation process, and its state vector includes the transportation resource occupation rate, the storage resource occupation rate, the task processing progress and the node load balancing degree, and the control vector includes the transportation resource adjustment amount and the storage resource adjustment amount. After linearization processing, the state space expression constructs a linear prediction model, and the task tracking prediction cost function includes a tracking error term, a control amount penalty term and a control increment penalty term. By solving the quadratic programming problem through the efficient set method, the optimal solution is gradually approached and the final execution control instruction is generated.

[0091] During the whole process, seamless connection is realized between modules through data interface, ensuring the accuracy and timeliness of information transmission. For example, in a certain scenario, when a batch of pallets need to complete the warehousing and delivery operation at the same time, the multi-source perception module first collects the identification, position and load state information of all pallets and generates a standardized state set. The task decomposition model generates a task level distribution according to the standardized state set, arranges the delivery task with high priority in front, and arranges the warehousing task with low priority in back. The multi-objective planning model calculates the optimal scheduling scheme through the dynamic allocation algorithm, ensuring the reasonable allocation of transportation resources and storage resources. The hierarchical execution control model generates execution control instructions according to the optimal scheduling scheme, guiding the logistics node to complete the specific warehousing and delivery operation. Through the above steps, the present application realizes efficient warehousing and delivery management of intelligent shared pallets, significantly improves the flow efficiency of pallets, and realizes efficient cooperation with the logistics system.

[0092] It should be noted that, in this paper, relational terms such as first and second are used merely to distinguish one entity or action from another, without necessarily requiring or implying that there is any such actual relationship or order between such entities or actions. Moreover, the terms "comprises", "comprising", or any other variations thereof are intended to cover non-exclusive inclusions, so that a process, method, article, or apparatus including a series of elements includes not only those elements, but also other elements not explicitly listed, or inherent to such a process, method, article, or apparatus.

[0093] Although embodiments of the present application have been shown and described, it will be understood by those having ordinary skill in the art that various changes, modifications, substitutions and alterations can be made hereto without departing from the principles and spirit of the present application, the scope of which is defined by the appended claims and their equivalents.

Claims

1. A smart shared pallet automated warehousing-in and warehousing-out method, characterized in that, include: The shared pallet's operational status data is collected through a multi-source sensing module, which includes a pallet identification unit, a pallet location tracking unit, and a pallet load monitoring unit. The operational status data of the shared pallet is preprocessed based on a hierarchical parsing framework to obtain a standardized status set. This standardized status set is then input into a pre-configured task deconstruction model, which employs a partitioned scheduling mechanism to hierarchically divide inbound and outbound tasks based on business logic rules, generating task deconstruction parameters. A multi-objective planning model is constructed based on the task deconstruction parameters. The multi-objective planning model aims to maximize the flow efficiency and minimize resource conflicts. A dynamic allocation algorithm is used to globally schedule inbound and outbound tasks. The dynamic allocation algorithm introduces task priority identification and resource suitability evaluation. The optimal scheduling scheme is output based on the multi-objective planning model. A hierarchical execution control model is established based on the optimal scheduling scheme. The hierarchical execution control model includes a strategy layer, a coordination layer, and an execution layer. The strategy layer performs global task planning based on the task decomposition parameters, the coordination layer performs local resource adjustment based on the optimal scheduling scheme, and the execution layer implements task tracking of logistics nodes based on a load balancing algorithm. The hierarchical execution control model outputs execution control commands to achieve efficient inbound and outbound management of shared pallets. The process involves inputting the standardized state set into a pre-configured task deconstruction model. This model employs a partitioned scheduling mechanism, hierarchically dividing inbound and outbound tasks based on business logic rules. The generated task deconstruction parameters include: obtaining a standardized state set, which includes pallet identification information, pallet location information, pallet load status, and cross-system association identifiers; constructing a state feature space based on the standardized state set, and constructing a resource capability space based on the allocatable transportation resources, storage capacity, and network communication bandwidth of the logistics platform; and constructing multi-dimensional business logic rules based on the state feature space and the resource capability space. These multi-dimensional business logic rules include task type matching items, processing timeliness requirements, resource adaptation items, and cross-system collaboration items. The task type matching items are obtained through compatibility evaluation of task attributes and processing modules, and the processing timeliness requirements are calculated through the matching degree between task update frequency and completion time. The resource adaptation items are obtained through task scale and resource load balancing evaluation, and the cross-system collaboration items are determined through the correspondence between association identifiers and collaborative processing flows. A partitioned scheduling mechanism is constructed, which includes an initial task set, hierarchical partitioning rules, and an association verification module. The initial task set contains multiple inbound and outbound task units. The hierarchical partitioning rules sort the task units hierarchically based on business process priorities. The association verification module verifies the dependencies between task units based on the multi-dimensional business logic rules. An iterative optimization method is used to update the initial task set, retaining high-priority task units through priority filtering, merging low-priority task units, and splitting complex task units. Based on the updated task set, the hierarchical distribution and resource requirements of inbound and outbound tasks are calculated. Based on the partitioned scheduling mechanism, task decomposition parameters are output, including task hierarchical distribution, resource requirement list, task unit dependencies, and cross-system collaboration identifiers. A multi-objective programming model is constructed based on the task deconstruction parameters. This model optimizes for maximizing flow efficiency and minimizing resource conflicts. A dynamic allocation algorithm is used for global scheduling of inbound and outbound tasks. The optimal scheduling scheme output based on the multi-objective programming model includes: constructing a multi-objective function for planning, comprising a flow efficiency optimization objective function and a resource conflict optimization objective function. The flow efficiency optimization objective function is calculated by summing the ratios of task unit processing time to corresponding resource processing speed. The resource conflict optimization objective function is calculated by weighted summing of the transportation resource occupancy rate, storage resource occupancy rate, and network communication occupancy rate of each logistics node. Planning constraints are constructed based on the multi-objective function, including time constraints, resource constraints, priority constraints, and coordination constraints. The time constraints limit the completion time range of inbound and outbound tasks, the resource constraints limit the resource usage upper limit of each logistics node, the priority constraints limit the scheduling order of high-priority task units, and the coordination constraints ensure the consistency of the processing order of task units across systems. A node-based approach is used. The sequence encodes inbound and outbound tasks, with each node containing a task unit identifier and corresponding logistics node information. A task allocation probability is constructed based on task transfer costs and heuristic evaluation values, and the next feasible node is determined using these probabilities. The heuristic evaluation values ​​are obtained through a weighted calculation of differences in processing time and resource consumption between nodes. An adaptive task transfer cost mechanism is introduced, employing a time-varying weight coefficient that linearly decreases with the number of scheduling iterations, dynamically adjusting the task transfer cost. A dynamic heuristic evaluation mechanism is also introduced, using a time-varying influence factor that linearly increases with the number of scheduling iterations, adjusting the weight of the heuristic evaluation value in the task allocation probability. Iterative optimization is performed based on the adaptive task transfer cost mechanism and the dynamic heuristic evaluation mechanism. The scheduling scheme generated in each iteration is evaluated for flow cost and resource cost, and non-dominated solutions are added to the Pareto front solution set. From the Pareto front solution set, solutions satisfying the flow cost-reduction principle are selected. The optimal solution for resource trade-offs is used as the optimal scheduling scheme. The optimal scheduling scheme is then processed to smooth the process and generate the optimal node sequence and the corresponding logistics node resource allocation parameters. The strategy layer performs global task planning based on the task deconstruction parameters, including: describing the task processing path using a parameterized process, representing the task processing path as a function of process parameters, where the process parameters range from 0 to 1, and the task processing path includes a task input stage, an intermediate processing stage, and a result output stage; describing the task processing path based on a piecewise linear function, calculating the time progress of the task processing path by summing the product of the control node coordinates and the segment weights, where the segment weights are calculated as the ratio of task unit priority to processing complexity; constructing task processing path constraints, including time constraints, sequence constraints, resource constraints, and quality constraints, where the time constraints limit the time progress range of the task processing path, the sequence constraints limit the execution order of task units, the resource constraints limit the upper limit of resource usage at each stage, and the quality constraints limit the accuracy requirements of the task processing results; and constructing a global multi-objective optimization function, which includes a total process duration term and resource usage fluctuation. The global multi-objective optimization function is weighted by a weighting factor, which includes a term, a sequence deviation term, and a quality compliance term. The task processing path interval is discretized into multiple process segments, and the global multi-objective optimization function is discretized to construct a global discretized objective function. The global discretized objective function includes process segment duration, resource usage, sequence deviation value, and quality compliance rate. The global discretized objective function is iteratively optimized using a gradient descent method. By calculating the gradient value of the objective function with respect to the control nodes, the time progress of the control nodes is updated along the negative gradient direction. The optimized task processing path is then smoothed by linear interpolation. By maintaining the continuity of time progress, resource usage, and sequence relationship at the interpolation endpoints, a smooth and continuous task processing path is generated. Based on the smooth and continuous task processing path, a node sequence and the corresponding logistics node processing speed are generated. The coordination layer performs local resource adjustments based on the optimal scheduling scheme, including: constructing a local adjustment window based on the current load, current resource occupancy, and current processing progress of the logistics platform; adaptively adjusting the size of the local adjustment window using a load correlation coefficient; establishing a positive correlation between the size of the local adjustment window and the current load; constructing a local resource model using real-time monitoring data; normalizing resource usage data to obtain standardized resource occupancy data; updating the availability flag of the resource status table based on the standardized resource occupancy data; calculating the resource availability probability value for each logistics node using a probability accumulation method; and predicting dynamic resource demand using an exponential smoothing algorithm, predicting the state vector of future resource demand using historical demand data, and updating the predicted state based on real-time monitoring data to obtain accurate resource demand. Predicting and analyzing trends; constructing a resource adjustment model, using the resource allocation rules of the logistics platform as the state equation, which includes transportation resources, storage resources, and network communication resources; constructing state constraints and allocation constraints, whereby the state constraints limit the range of resource occupancy and processing progress, and the allocation constraints limit the range of resource allocation adjustment magnitude and frequency; constructing a multi-objective cost function, which includes reference scheduling tracking, resource conflict avoidance, process smoothing, and energy consumption, and weighting the terms of the multi-objective cost function using weighting coefficients; optimizing the multi-objective cost function using the Lagrange multiplier method, constructing a Lagrange function and introducing constraints, and obtaining the optimal adjustment amount by solving the partial derivative equations; The execution layer implements task tracking for logistics nodes based on a load balancing algorithm. Through the hierarchical execution control model, it outputs execution control commands to achieve efficient inbound and outbound management of shared pallets. This includes: establishing a multi-node resource allocation model for the logistics platform, comprising a transportation resource allocation equation and a storage resource allocation equation. The transportation resource allocation equation includes the processing workload, node transportation capacity, and task priority. The storage resource allocation equation includes data storage volume, node storage capacity, and data access frequency. The multi-node resource allocation model is constructed as a state-space expression. The state vector of the state-space expression includes transportation resource occupancy rate, storage resource occupancy rate, task processing progress, and node load balancing degree. The control vector of the state-space expression includes transportation resource adjustment and storage resource adjustment. The state-space expression is linearized, and the partial derivatives of the system state equation with respect to the state vector and control vector are calculated to construct a linearized prediction model. A task tracking prediction cost function is constructed, comprising a tracking error term, a control quantity penalty term, and a control increment penalty term. This is achieved through weighting... The system employs a weighted combination of penalty terms using a multiplicative matrix; it constructs state constraints, including resource occupancy constraints and task processing progress constraints; it also constructs control constraints, including transportation resource adjustment constraints and storage resource adjustment constraints; and it constructs control increment constraints, including transportation resource increment constraints and storage resource increment constraints. The task tracking prediction cost function is transformed into the standard form of a quadratic programming problem, and the quadratic form matrix and linear coefficients are calculated to construct inequality constraint matrices and equality constraint matrices. The system uses the effective set method to solve the quadratic programming problem, gradually approximating the optimal solution by identifying effective constraints and solving subproblems. Based on the optimized solution results, control quantities are mapped: the total transportation resource adjustment is allocated to each logistics node using the transportation resource allocation matrix, and the total storage resource adjustment is allocated to each node's storage unit using the storage resource allocation matrix. Resource allocation to logistics nodes is limited based on the node's maximum transportation capacity and storage resource allocation, generating the final execution control instructions. Based on the hierarchical parsing framework, the shared pallet's operational status data is preprocessed to obtain a standardized status set. This includes: converting the shared pallet operational status data collected by the pallet identification unit, pallet location tracking unit, and pallet load monitoring unit to obtain a unified data format; cleaning outliers from each format by calculating the mean and variance of the data using a sliding window to remove outliers that exceed the range; extracting the core business information of each data point using a keyword matching algorithm and establishing correspondences between data from different systems through information association; converting the associated business information into a unified data model and fusing data from multiple systems using a weighted average method to generate a standardized status set containing timestamps.

2. The intelligent shared pallet automated warehousing method according to claim 1, wherein, The partitioned scheduling mechanism includes: initializing the task set with uniform distribution, where each task unit contains the business type and time requirements of inbound and outbound tasks; calculating the node resource requirements corresponding to the task unit through the logistics platform's resource allocation rules, mapping the task unit requirements to the resource capability table based on interface parameter configuration, and generating a task unit resource requirement list; calculating the matching error between the task unit resource requirements and the actual node resources, and constructing business logic rules based on the sum of squared errors; obtaining the priority value of each task unit through normalization, retaining task units with priority values ​​greater than a preset threshold, and removing task units with priority values ​​less than a preset threshold.

3. The intelligent shared pallet automated warehousing method according to claim 1, wherein, The task transfer cost is constructed by: calculating the processing time difference between the current node and the next node as the processing time cost; calculating the adjustment changes of each resource required by the logistics platform from the current node to the next node as the resource adjustment cost; calculating the minimum distance of resource conflict in the scheduling scheme, and increasing the penalty cost if it is less than the safety threshold; and generating the total task transfer cost by linearly combining the processing time cost, resource adjustment cost, and penalty cost.

4. The intelligent shared pallet automated warehousing method according to claim 1, wherein, Solving the quadratic programming problem using the effective set method includes: initializing the effective constraint set and adding all constraints to the candidate set; solving the initial subproblem to obtain the initial feasible solution; calculating the constraint violation amount of the current solution and adding the constraint with the largest violation amount to the effective set; resolving the subproblem and updating the current solution; repeating the above steps until the violation amount of all constraints is less than the preset precision, thus obtaining the optimal solution.

Images