CN-121724553-B - Intelligent warehouse logistics real-time goods space optimization method and system

CN121724553BCN 121724553 BCN121724553 BCN 121724553BCN-121724553-B

Abstract

The application discloses a real-time cargo space optimizing method and system of intelligent warehouse logistics, relating to the technical field of intelligent warehouse logistics, by acquiring the real-time data of the warehouse, constructing a warehouse dynamic state map, generating an initial allocation scheme, inputting a distributed decision network for parallel collaborative evaluation optimization and issuing global instructions, the problems of delay and conflict of the traditional method in peak period are solved, and warehouse operation efficiency and system throughput can be improved.

Inventors

DU LIFENG
NIE FANG
YANG CHUNQING

Assignees

成都锦城学院

Dates

Publication Date: 20260508
Application Date: 20260226

Claims (8)

1. The intelligent warehouse logistics real-time goods position optimizing method is characterized by comprising the following steps: Acquiring storage real-time data, wherein the storage real-time data comprises real-time goods space state data, real-time handling equipment position data and real-time order stream data; The warehouse dynamic state map is used for representing the dynamic relationship between the task generated by the real-time order stream data and the position of the carrying equipment and the space occupation state at the current moment in a correlation graph; Determining a to-be-allocated cargo space set according to the real-time order stream data, and generating initial allocation scheme data for each to-be-allocated cargo space in the to-be-allocated cargo space set, wherein the initial allocation scheme data comprises candidate carrying equipment and an estimated execution path; Inputting the warehouse dynamic state map and all the initial allocation scheme data into a distributed decision network; The method comprises the steps of carrying out parallelization collaborative evaluation and optimization on all initial allocation scheme data through a distributed decision network to generate a global optimization instruction set, wherein the collaborative evaluation is based on a dynamic conflict prediction mechanism, and the dynamic conflict prediction mechanism analyzes the conflict possibility of execution of different initial allocation schemes under the same space-time resource by analyzing the overlapping condition of estimated execution paths of different transport tasks on the future space-time dimension; issuing the global optimization instruction set to corresponding carrying equipment so as to drive the carrying equipment to execute a real-time goods space optimization task; through the distributed decision network, performing parallelization collaborative evaluation and optimization on all the initial allocation scheme data, and generating a global optimization instruction set comprises the following steps: Distributing the initial distribution scheme data to at least one corresponding sub-decision-making intelligent agent according to the storage area covered by the estimated execution path, wherein the at least one sub-decision-making intelligent agent corresponds to different storage areas and forms a distributed decision-making network; each sub-decision agent performs local conflict previewing based on the received initial allocation scheme data and local map data related to the local area in the warehouse dynamic state map to generate local evaluation result data, wherein the local evaluation result data comprises a path collision risk value and expected completion time; All the sub-decision agents send the respective local evaluation result data to a central coordinator; the central coordinator integrates all the local evaluation result data, arbitrates and adjusts according to a global optimization target, and generates the global optimization instruction set; Each sub-decision agent performs local conflict previewing based on the received initial allocation scheme data and local map data related to the local area in the warehouse dynamic state map, and the step of generating local evaluation result data comprises the following steps: Based on the local map data, simulating and predicting the occupation condition of the current region in a preset future time window as space-time occupation thermodynamic map data; Obtaining an estimated execution path from the initial allocation scheme data, mapping the estimated execution path onto the space-time occupation thermodynamic diagram data according to the execution time sequence, and calculating the space-time overlapping degree of the path to obtain path collision risk assessment data; Analyzing the attribute information of the handling equipment and the real-time state information thereof contained in the local map data, acquiring the performance parameters of the handling equipment and the current task queue state corresponding to the initial allocation scheme data, and calculating expected completion time data based on the length and the complexity of the estimated execution path; And combining the path collision risk assessment data, the expected completion time data and the emergency degree label in the task data corresponding to the initial allocation scheme data, and calculating and integrating by adopting a preset weighting scoring rule to obtain the local assessment result data comprising the path collision risk value and the expected completion time.
2. The intelligent warehouse logistics real-time cargo space optimization method of claim 1, wherein the step of constructing a warehouse dynamic state map based on the warehouse real-time data comprises: performing gridding encoding processing on the real-time cargo space state data to obtain cargo space grid encoding data; performing coordinate mapping processing on the real-time handling equipment position data to obtain equipment grid coordinate data of the same coordinate system as the cargo space grid coding data; Performing task analysis processing on the real-time order stream data to obtain a corresponding task data set, wherein the task data at least comprises a target goods position identifier and an emergency degree label; Generating a node set and an edge set based on the cargo space grid coding data, the equipment grid coordinate data and the task data set, wherein nodes in the node set are used for representing cargo space, carrying equipment and task entities, and edges in the edge set are used for representing spatial adjacent relations, task subordinate relations and equipment load relations among the nodes; And constructing the warehouse dynamic state map according to the node set and the edge set.
3. The intelligent warehouse logistics real-time cargo space optimization method as claimed in claim 2, wherein the task parsing process comprises: Analyzing commodity characteristic information in the real-time order stream data to obtain characteristic data comprising commodity size and commodity weight; inquiring candidate cargo space specification data which is matched with the characteristic data and meets the preset physical safety margin from a pre-stored cargo space specification database, wherein the cargo space specification data comprises a cargo space bearing upper limit and a cargo space size; According to the candidate cargo space specification data, combining current vacant cargo space information in the warehouse real-time data to perform dynamic compatibility verification, namely comparing the commodity size and the commodity weight with the space size and the bearing upper limit of the vacant cargo space respectively, and screening cargo spaces with bearing capacity or space allowance lower than a preset safety threshold to generate physical compatible cargo space inventory data; and adding the physical compatible goods-space inventory data serving as the goods-space physical attribute requirement data to corresponding task data so as to guide screening of the goods-space set to be distributed.
4. The method of optimizing real-time cargo space of intelligent warehouse logistics according to claim 1, wherein the step of simulating and predicting the occupation of the current area in a preset future time window as space-time occupation thermodynamic diagram data based on the local map data comprises: Predicting the expected position of the carrying equipment at each moment in the preset future time window according to the current speed, the current position and the known task path of the carrying equipment in the local map data to form equipment expected track data; Predicting the change of the occupancy state of the cargo space in the preset future time window according to the current occupancy state of the cargo space in the local map data and the known time of the to-be-executed loading or unloading task to form cargo space occupancy prediction data; And fusing the expected track data of the equipment and the goods space occupation prediction data to generate the space-time occupation thermodynamic diagram data taking grids as units and taking time as dimensions, wherein the value of each grid unit represents the probability that the unit is occupied at the corresponding time.
5. The intelligent warehouse logistics real-time cargo space optimization method of claim 1, wherein the step of the central coordinator integrating all the local evaluation result data and arbitrating and adjusting according to a global optimization objective to generate the global optimization instruction set comprises: Receiving all the local evaluation result data, and identifying paired or grouped initial allocation scheme data in which the path collision risk value exceeds a preset risk threshold value; For the identified initial allocation scheme data, starting a priority arbitration flow, namely comparing urgency labels or task types of orders associated with tasks of conflict parties, giving different execution priorities according to preset arbitration rules, and generating a priority arbitration result; Based on the priority arbitration result, carrying out dynamic re-planning on the estimated execution path corresponding to the initial allocation scheme data judged to be low in priority, generating re-planned path data, or directly generating equipment waiting instructions for instructing the corresponding carrying equipment to execute waiting operation; And packaging all the arbitrated and adjusted task execution schemes, the re-planning path data and the equipment waiting instruction into the global optimization instruction set.
6. The intelligent warehouse logistics real-time cargo space optimization method of claim 1, wherein the step of parallelizing collaborative evaluation and optimization of all the initial allocation scheme data through the distributed decision network, generating a global optimization instruction set further comprises: periodically counting the received task data set to obtain the newly added task data amount in unit time, and calculating the ratio of the task data amount to the average task processing completion rate of the system to obtain the real-time load rate data of the system; Monitoring the average task queue length of all the carrying devices, and generating device queue pressure data; When the real-time load rate data of the system continuously exceeds a first threshold value for a preset duration and the pressure data of the equipment queue simultaneously exceeds a second threshold value, judging to enter a high load state; Under the high-load state, screening out high-value tasks according to preset high-value attribute labels of the tasks in the task data set, and generating a high-value task set; Suspending the generation and optimization processing of the initial allocation scheme of the tasks in the non-high-value task set, focusing the evaluation and optimization resources of the distributed decision network on the initial allocation scheme data corresponding to the high-value task set, and rapidly generating and issuing the global optimization instruction set for the high-value task.
7. The method for optimizing real-time cargo space of intelligent warehouse logistics according to claim 6, wherein the steps of screening out high-value tasks according to the preset high-value attribute labels of the tasks in the task data set and generating a high-value task set comprise: Setting weight coefficients for different preset high-value attribute tags, and constructing a multidimensional value evaluation model; Extracting the numerical value of the corresponding preset high-value attribute label for each task; Weighting and calculating the numerical value by using the multi-dimensional value evaluation model to obtain the comprehensive value score of the task; And judging the task with the comprehensive value score higher than a dynamic value score threshold as a high-value task and collecting the high-value task set, wherein the dynamic value score threshold is adjusted in real time according to the current system real-time load rate data and the equipment queue pressure data.
8. An intelligent warehouse logistics real-time cargo space optimizing system, characterized in that the intelligent warehouse logistics real-time cargo space optimizing system comprises a memory, a processor and an intelligent warehouse logistics real-time cargo space optimizing program which is stored on the memory and can run on the processor, wherein the intelligent warehouse logistics real-time cargo space optimizing program is configured to realize the steps of the intelligent warehouse logistics real-time cargo space optimizing method according to any one of claims 1 to 7.

Description

Intelligent warehouse logistics real-time goods space optimization method and system Technical Field The application relates to the technical field of intelligent warehouse logistics, in particular to a real-time goods space optimizing method and system for intelligent warehouse logistics. Background In the field of intelligent warehouse logistics, real-time goods space optimization is used as a core link for improving warehouse operation efficiency, and order processing speed and resource utilization rate are directly affected. The traditional warehouse management system generally adopts a static path planning algorithm or a goods space allocation strategy based on historical order data, and the method can maintain basic efficiency under the conventional operation state of stable order flow and small change of the working environment. However, with the rapid development of electronic commerce, the warehouse work environment presents dynamic and high concurrency characteristics, especially during large-scale sales promotion, the order volume is often multiplied by several times in a short time, forming an extreme peak scenario. In the prior art, multiple challenges are faced under the scene that firstly, an optimization decision highly depends on a centralized computing architecture of a central server, when a massive real-time order stream is faced, a system is difficult to complete quick analysis, task decomposition and preliminary allocation of orders within a millisecond time window, so that tasks are generated to be accumulated continuously, secondly, after an initial carrying scheme is generated, the system lacks collaborative prediction capability of potential conflict between a plurality of carrying devices and a plurality of tasks in a future space-time dimension, risks such as path crossing, device collision or resource contention cannot be effectively identified, thirdly, due to the lack of a dynamic arbitration mechanism, when the plurality of tasks generate resource competition in a shared storage space, the system cannot conduct real-time priority adjustment according to the emergency degree or the state of the tasks, and further the problems of frequent waiting of the carrying devices, low path re-planning efficiency and the like are caused. The defects directly cause the great prolongation of the response time of the system in the peak period, the frequent occurrence of equipment path conflict events and the reduction of the overall throughput of warehousing operation, restrict the timeliness and the reliability of logistics service, and are difficult to meet the requirements of modern intelligent logistics on real-time response, high concurrency processing and system robustness. The foregoing is provided merely for the purpose of facilitating understanding of the technical solutions of the present application and is not intended to represent an admission that the foregoing is prior art. Disclosure of Invention The application mainly aims to provide an intelligent warehouse logistics real-time cargo space optimizing method and system, aiming at improving the response speed of warehouse operation, reducing the equipment conflict event and improving the overall throughput of a system. In order to achieve the above purpose, the application provides a real-time cargo space optimizing method for intelligent warehouse logistics, which comprises the following steps: Acquiring storage real-time data, wherein the storage real-time data comprises real-time goods space state data, real-time handling equipment position data and real-time order stream data; The warehouse dynamic state map is used for representing the dynamic relationship between the task generated by the real-time order stream data and the position of the carrying equipment and the space occupation state at the current moment in a correlation graph; Determining a to-be-allocated cargo space set according to the real-time order stream data, and generating initial allocation scheme data for each to-be-allocated cargo space in the to-be-allocated cargo space set, wherein the initial allocation scheme data comprises candidate carrying equipment and an estimated execution path; Inputting the warehouse dynamic state map and all the initial allocation scheme data into a distributed decision network; Performing parallelization collaborative evaluation and optimization on all the initial allocation scheme data through the distributed decision network to generate a global optimization instruction set, wherein the collaborative evaluation is based on a dynamic conflict prediction mechanism, and the dynamic conflict prediction mechanism is used for analyzing the conflict possibility of executing different initial allocation schemes under the same space-time resource; And issuing the global optimization instruction set to corresponding carrying equipment so as to drive the carrying equipment to execute a real-time goods space optimization task. In an