Storage strategy optimization method and device, electronic equipment and storage medium

Abstract

The invention provides a storage strategy optimization method and device, electronic equipment and a storage medium, and the method comprises the steps: obtaining an initial storage strategy, inputting the initial storage strategy into a preset storage simulation model, and obtaining initial index values under a plurality of preset dimensions, obtaining a strategy score of the initial storage strategy based on the initial index value and a corresponding preset weight value, inputting the initial index value, the initial storage strategy and the corresponding strategy score into a large language model to generate a candidate storage strategy, taking the candidate storage strategy as an iterative storage strategy, and repeatedly executing an iterative operation to obtain an iterative storage strategy; and taking the iterative warehousing strategy as an optimized target warehousing strategy until the strategy score of the iterative warehousing strategy is greater than a preset score threshold value, thereby solving the technical problems of single warehousing strategy evaluation dimension and insufficient optimization capability in related technologies.

Description

Technical Field

[0001] This invention relates to the field of heavy-duty warehousing technology, specifically to a warehousing strategy optimization method, apparatus, electronic device, and storage medium. Background Technology

[0002] As a fundamental industry of the national economy, the continuity and efficiency of the metallurgical manufacturing process are crucial. In the complex material flow system of metallurgical production, warehousing and logistics, as the "throat" connecting various processes, directly determine the cost, response speed, and flexibility of the entire production chain. With the deepening of intelligent manufacturing, metallurgical warehousing systems are developing towards high automation and intelligence. However, how to scientifically and comprehensively evaluate the efficiency of this highly complex system and achieve continuous and precise optimization based on this evaluation has become a core challenge facing the industry.

[0003] Currently, there are significant limitations in efficiency optimization technologies in the metallurgical warehousing field. First, at the efficiency evaluation level, these technologies often rely on isolated, single, local performance indicators. Second, at the optimization strategy generation level, these methods often depend on pre-set fixed rules or experts' historical experience, leading to optimization results that frequently get stuck in local optima and fail to fully tap the system's potential efficiency. This results in recurring efficiency bottlenecks such as process conflicts, resource idleness, and slow response times. Summary of the Invention

[0004] This invention provides a method, apparatus, electronic device, and storage medium for optimizing warehousing strategies, in order to solve the technical problems of insufficient optimization capabilities and limited evaluation dimensions for warehousing strategies in related technologies.

[0005] This invention provides a warehouse strategy optimization method, the method comprising: obtaining an initial warehouse strategy; inputting the initial warehouse strategy into a preset warehouse simulation model to obtain initial indicator values ​​under multiple preset dimensions; obtaining a strategy score of the initial warehouse strategy based on the initial indicator values ​​and their corresponding preset weight values; inputting the initial indicator values, the initial warehouse strategy, and the corresponding strategy score into a large language model to generate candidate warehouse strategies; and using the candidate warehouse strategies as iterative warehouse strategies to perform an iterative operation, the iterative operation comprising: inputting the iterative warehouse strategy into the warehouse simulation model to obtain... Based on the current indicator value under the multiple preset dimensions and its corresponding preset weight value, the strategy score of the iterative storage strategy is obtained. If the strategy score of the iterative storage strategy is less than a preset score threshold, the current indicator value, the iterative storage strategy, and the corresponding strategy score are input into the large language model again to generate a new candidate storage strategy, and the new candidate storage strategy is used as the updated iterative storage strategy. The iterative operation is repeated until the strategy score of the iterative storage strategy is greater than the preset score threshold, and the iterative storage strategy is used as the optimized target storage strategy.

[0006] In one embodiment of the present invention, inputting the initial warehousing strategy into a preset warehousing simulation model to obtain initial index values ​​under multiple preset dimensions includes: inputting the process parameters and equipment configuration corresponding to the initial warehousing strategy into the warehousing simulation model to obtain initial index values ​​under multiple preset dimensions, wherein the warehousing simulation model includes a digital twin model based on discrete event simulation; wherein the process parameters include at least one of inbound material characteristics, outbound order patterns, and storage strategy rules, and the equipment configuration includes at least one of warehouse layout structure, handling equipment performance parameters, and storage equipment attributes.

[0007] In one embodiment of the present invention, obtaining the strategy score of the initial warehousing strategy based on the initial indicator value and its corresponding preset weight value includes: normalizing the initial indicator value; and performing a weighted summation based on the normalized initial indicator value and its corresponding preset weight value to obtain the strategy score of the initial warehousing strategy.

[0008] In one embodiment of the present invention, before inputting the initial indicator value, the initial warehousing strategy, and the corresponding strategy score into the large language model, the method further includes: acquiring multi-source warehousing information, wherein the multi-source warehousing information includes at least historical warehousing operation data, professional documents in the warehousing field, and warehousing industry knowledge; vectorizing the multi-source warehousing information based on text embedding technology to obtain text vectors; and mapping and associating the text vectors with entities in the preset knowledge graph of the large language model to enhance the knowledge of the large language model.

[0009] In one embodiment of the present invention, inputting the initial indicator value, the initial storage strategy, and the corresponding strategy score into a large language model to generate a candidate storage strategy includes: constructing a prompt word including an optimization instruction, wherein the prompt word further includes the initial indicator value, the initial storage strategy, and the corresponding strategy score, the optimization instruction including an instruction for guiding the large language model to optimize the initial indicator value under the multiple preset dimensions, and inputting the prompt word into the large language model to obtain at least one candidate storage strategy generated by the large language model based on the prompt word.

[0010] In one embodiment of the present invention, if there are multiple candidate warehousing strategies, then using the candidate warehousing strategies as iterative warehousing strategies includes: inputting the multiple candidate warehousing strategies into the warehousing simulation model to obtain candidate index values ​​under multiple preset dimensions corresponding to the multiple candidate warehousing strategies; comparing the candidate index values ​​under the multiple preset dimensions corresponding to the multiple candidate warehousing strategies pairwise, and determining the iterative warehousing strategy based on the comparison results.

[0011] In one embodiment of the present invention, after repeatedly executing the iterative operation, the method further includes: if the number of times the iterative operation is repeatedly executed exceeds a preset threshold, then the repeated execution of the iterative operation is stopped, and the iterative storage strategy with the highest strategy score is determined as the target storage strategy.

[0012] The present invention also provides a warehousing strategy optimization device, the device comprising: an initial strategy input module, configured to acquire an initial warehousing strategy, input the initial warehousing strategy into a preset warehousing simulation model, and obtain initial index values ​​under multiple preset dimensions; an initial strategy scoring module, configured to obtain a strategy score of the initial warehousing strategy based on the initial index values ​​and their corresponding preset weight values; a candidate strategy generation module, configured to input the initial index values, the initial warehousing strategy, and the corresponding strategy score into a large language model to generate candidate warehousing strategies; and a candidate strategy iteration module, configured to use the candidate warehousing strategies as iterative warehousing strategies and perform iterative operations, the iterative operations including: optimizing the iterative warehousing strategy... The strategy is input into the warehouse simulation model to obtain the current indicator value under the multiple preset dimensions. Based on the current indicator value and its corresponding preset weight value, the strategy score of the iterative warehouse strategy is obtained. If the strategy score of the iterative warehouse strategy is less than the preset score threshold, the current indicator value, the iterative warehouse strategy and the corresponding strategy score are input into the large language model again to generate a new candidate warehouse strategy, and the new candidate warehouse strategy is used as the updated iterative warehouse strategy. The target strategy optimization module is used to repeatedly execute the iterative operation until the strategy score of the iterative warehouse strategy is greater than the preset score threshold, and the iterative warehouse strategy is used as the optimized target warehouse strategy.

[0013] The present invention also provides an electronic device, comprising: one or more processors; and a storage device for storing one or more programs, wherein when the one or more programs are executed by the one or more processors, the electronic device enables the warehousing strategy optimization method as described in any of the above embodiments.

[0014] The present invention also provides a computer-readable storage medium storing a computer program, which, when executed by a computer's processor, causes the computer to perform any of the warehouse strategy optimization methods described in the above embodiments.

[0015] The beneficial effects of this invention are as follows: This invention proposes a storage strategy optimization method, apparatus, electronic device, and storage medium. By acquiring an initial storage strategy and inputting it into a preset storage simulation model, initial index values ​​are obtained under multiple preset dimensions. Based on the initial index values ​​and their corresponding preset weight values, a strategy score for the initial storage strategy is obtained. The initial index values, the initial storage strategy, and the corresponding strategy score are input into a large language model to generate candidate storage strategies. The candidate storage strategies are used as iterative storage strategies, and iterative operations are repeatedly performed until the strategy score of the iterative storage strategy exceeds a preset score threshold. The iterative storage strategy is then used as the optimized target storage strategy. This invention, through a more comprehensive and systematic strategy evaluation scheme and the introduction of a large language model to generate diverse candidate strategies, can achieve more intelligent strategy optimization, thereby improving the overall efficiency of the storage system.

[0016] It should be understood that the above general description and the following detailed description are exemplary and explanatory only, and are not intended to limit the invention. Attached Figure Description

[0017] The accompanying drawings, which are incorporated in and constitute a part of this specification, illustrate embodiments consistent with the invention and, together with the description, serve to explain the principles of the invention. It is obvious that the drawings described below are merely some embodiments of the invention, and those skilled in the art can obtain other drawings based on these drawings without any inventive effort.

[0018] In the attached diagram: Figure 1 This is a schematic diagram illustrating the implementation environment of a warehousing strategy optimization method according to an embodiment of the present invention. Figure 2 This is a flowchart of a warehousing strategy optimization method provided in one embodiment of the present invention; Figure 3 This is a block diagram of a warehousing strategy optimization device provided in one embodiment of the present invention; Figure 4This is a schematic diagram of the structure of an electronic device provided in one embodiment of the present invention. Detailed Implementation

[0019] The following specific examples illustrate the implementation of the present invention. Those skilled in the art can easily understand other advantages and effects of the present invention from the content disclosed in this specification. The present invention can also be implemented or applied through other different specific embodiments. Various details in this specification can also be modified or changed based on different viewpoints and applications without departing from the spirit of the present invention. In the absence of conflict, the following embodiments and features in the embodiments can be combined with each other.

[0020] It should be noted that the illustrations provided in the following embodiments are only schematic representations of the basic concept of the present invention. The drawings only show the components related to the present invention and are not drawn according to the actual number, shape and size of the components in the actual implementation. In the actual implementation, the form, quantity and proportion of each component can be arbitrarily changed, and the layout of the components may also be more complex.

[0021] In the following description, numerous details are explored to provide a more thorough explanation of embodiments of the invention. However, it will be apparent to those skilled in the art that embodiments of the invention may be practiced without these specific details. In other embodiments, well-known structures and devices are shown in block diagram form rather than in detail to avoid obscuring embodiments of the invention.

[0022] Please see Figure 1 , Figure 1 This is a schematic diagram illustrating the implementation environment of a warehousing strategy optimization method according to an embodiment of the present invention.

[0023] like Figure 1As shown, the implementation environment may include computer equipment 110, which can be at least one of microcomputers, embedded computers, neural network computers, etc. Computer equipment 110 is equipped with at least a warehouse simulation platform, a large language model, a vector database and retrieval module, a knowledge graph management system, and a multi-objective optimization algorithm module. The warehouse simulation platform is a software environment capable of building and running discrete event simulation models (warehouse simulation models), such as a self-developed simulation engine based on languages ​​like Python (SimPy) or Java. This platform is responsible for executing the simulation of the strategy and outputting multi-dimensional initial and candidate indicator values. The large language model serves as the "intelligent engine" for strategy generation. This can be achieved by calling cloud APIs of other intelligent models or by deploying open-source or self-developed large models (such as LLaMA and ChatGLM series) on a local server to ensure data security and customization. The vector database and retrieval module is used to achieve vectorized storage and fast similarity retrieval of unstructured text knowledge, for example, using vector database technologies such as Milvus, Pinecone, and Chroma. Knowledge graph management systems are used to build, store, and query structured domain knowledge graphs, such as those using graph databases like Neo4j and Nebula Graph. The multi-objective optimization algorithm module includes a program with built-in Pareto selection logic to automatically identify the set of non-dominated policies from multiple candidate policies.

[0024] Please see Figure 2 , Figure 2 This is a flowchart illustrating a warehousing strategy optimization method provided in one embodiment of the present invention. This method can be applied to... Figure 1 The implementation environment shown can also be applied to other exemplary implementation environments and specifically executed by devices in other implementation environments. This embodiment does not limit the implementation environment to which the method is applicable.

[0025] like Figure 2 As shown, in an exemplary embodiment, the warehousing strategy optimization method includes at least steps S210 to S250, which are described in detail below: Step S210: Obtain the initial warehousing strategy and input the initial warehousing strategy into the preset warehousing simulation model to obtain the initial index values ​​under multiple preset dimensions.

[0026] For example, the purpose of the initial indicator values ​​is to measure the efficiency of warehousing under different warehousing strategies, including the initial indicator values ​​of multiple secondary indicators set under multiple preset dimensions.

[0027] In one embodiment of the present invention, inputting the initial warehousing strategy into a preset warehousing simulation model to obtain initial index values ​​under multiple preset dimensions includes: inputting the process parameters and equipment configuration corresponding to the initial warehousing strategy into the warehousing simulation model to obtain initial index values ​​under multiple preset dimensions. The warehousing simulation model includes a digital twin model based on discrete event simulation. The process parameters include at least one of the following: characteristics of inbound materials, outbound order patterns, and storage strategy rules. The equipment configuration includes at least one of the following: warehouse layout structure, handling equipment performance parameters, and storage equipment attributes.

[0028] For example, the preset multiple dimensions may include a process efficiency dimension, a spatial configuration dimension, an equipment operation dimension, and a scheduling decision dimension. Each dimension can have multiple secondary indicators. For instance, under the process efficiency dimension, indicators for measuring process standardization and the percentage of time spent on bottleneck stages can be set. These secondary indicators are divided into positive and negative indicators. A higher value for a positive indicator indicates better performance, while a lower value indicates better performance. For example, the process standardization indicator is a positive indicator, and the percentage of time spent on bottleneck stages is a negative indicator. In this embodiment, the initial indicator values ​​are the simulation output data obtained after simulating the initial warehousing strategy in the warehousing simulation model, and the indicator values ​​of multiple secondary indicators under the four dimensions are calculated based on this simulation output data.

[0029] For example, the preset dimensions and secondary indicators can be customized according to the actual warehousing strategy. In this embodiment, the process efficiency dimension may include secondary indicators related to operational coordination efficiency and bottleneck time consumption. The space configuration efficiency dimension may include secondary indicators related to the rationality of warehouse layout and storage utilization. The equipment operation efficiency dimension may include secondary indicators related to equipment overall utilization and operating cycle time. The scheduling decision efficiency dimension may include secondary indicators related to task allocation, path planning, and conflict resolution efficiency.

[0030] Step S220: Obtain the strategy score of the initial warehousing strategy based on the initial indicator value and its corresponding preset weight value.

[0031] In one embodiment of the present invention, obtaining the strategy score of the initial warehousing strategy based on the initial index value and its corresponding preset weight value includes: normalizing the initial index value; and performing a weighted summation based on the normalized initial index value and its corresponding preset weight value to obtain the strategy score of the initial warehousing strategy.

[0032] For example, the preset weight values ​​include dimension weights and indicator weights. The preset weight values ​​can be set using the analytic hierarchy process and the preset weight values ​​can be validated for consistency, thereby reasonably setting dimension weights for each dimension and indicator weights for each initial indicator value.

[0033] For example, normalizing the initial indicator value includes normalizing it according to the following formula if the initial indicator value is a positive indicator: Equation (1) In equation (1), The initial index value after normalization. This is the initial index value.

[0034] For example, normalizing the initial indicator value includes normalizing it according to the following formula if the initial indicator value is an inverse indicator: Equation (2) In equation (2), The initial index value after normalization. This is the initial index value.

[0035] For example, the strategy score of the initial warehousing strategy, based on the initial indicator values ​​and their corresponding preset weight values, includes: Equation (3) In equation (3), P is the strategy score, and i is the i-th preset dimension. Let be the dimension weight corresponding to the i-th dimension, and j be the j-th secondary indicator under each preset dimension. Let the index weight be the index weight corresponding to the j-th secondary index under the i-th preset dimension. This is the normalized initial index value corresponding to the j-th secondary index under the i-th preset dimension.

[0036] Step S230: Input the initial indicator value, the initial storage strategy and the corresponding strategy score into the large language model to generate candidate storage strategies.

[0037] In one embodiment of the present invention, before inputting the initial indicator value, the initial warehousing strategy, and the corresponding strategy score into the large language model, the method further includes: acquiring multi-source warehousing information, which includes at least historical warehousing operation data, professional documents in the warehousing field, and warehousing industry knowledge; vectorizing the multi-source warehousing information based on text embedding technology to obtain text vectors; and mapping and associating the text vectors with entities in the preset knowledge graph of the large language model to enhance the knowledge of the large language model.

[0038] For example, the pre-defined knowledge graph includes entities and relationships in the warehousing domain. Entities include physical assets, operational objects, processes and rules, efficiency evaluation related concepts, etc. Physical assets include warehouse areas, storage locations, and overhead cranes, etc. Operational objects include materials such as steel coils and aluminum coils, processes and rules include warehousing processes, scheduling rules, and failure modes, etc. Efficiency evaluation related concepts include probabilistic events and indicator snapshots, etc. Relationships mainly define spatial relationships (e.g., "located in"), execution relationships (e.g., overhead crane "executes" instructions), and causal relationships (e.g., bottlenecks "affect" efficiency), etc., between the above entities. By vectorizing multi-source warehousing information to obtain text vectors, and mapping and associating these text vectors with entities in the knowledge graph, the knowledge graph instances are populated to enhance the knowledge of the large language model.

[0039] In one embodiment of the present invention, inputting initial indicator values, initial storage strategies, and corresponding strategy scores into a large language model to generate candidate storage strategies includes: constructing prompt words that include optimization instructions, wherein the prompt words also include initial indicator values, initial storage strategies, and corresponding strategy scores, and the optimization instructions include instructions for guiding the large language model to optimize initial indicator values ​​under multiple preset dimensions; inputting the prompt words into the large language model to obtain at least one candidate storage strategy generated by the large language model based on the prompt words.

[0040] For example, the large language model first generates a preliminary analysis request based on prompts including optimization instructions, initial metric values, initial storage strategies, and corresponding strategy scores. Then, it leverages the structured advantages of knowledge graphs to understand and expand the analysis request, and retrieves the most relevant contextual information, such as similar historical cases, related operating procedures, or potential causal chains. Based on the contextual information, the large language model performs in-depth analysis of the analysis request and generates a set of candidate storage strategies. These candidate storage strategies include diverse strategies generated based on multiple optimization objectives (such as improving utilization and reducing conflict rates).

[0041] Step S240: The candidate storage strategy is used as the iterative storage strategy, and an iterative operation is performed. The iterative operation includes: inputting the iterative storage strategy into the storage simulation model to obtain the current indicator value under multiple preset dimensions; obtaining the strategy score of the iterative storage strategy based on the current indicator value and its corresponding preset weight value; if the strategy score of the iterative storage strategy is less than the preset score threshold, the current indicator value, the iterative storage strategy and the corresponding strategy score are input into the large language model again to generate a new candidate storage strategy, and the new candidate storage strategy is used as the updated iterative storage strategy.

[0042] In one embodiment of the present invention, if there are multiple candidate warehousing strategies, the selection of candidate warehousing strategies as iterative warehousing strategies includes: inputting multiple candidate warehousing strategies into a warehousing simulation model to obtain candidate index values ​​under multiple preset dimensions corresponding to the multiple candidate warehousing strategies; comparing the candidate index values ​​under multiple preset dimensions corresponding to the multiple candidate warehousing strategies pairwise, and determining the iterative warehousing strategy based on the comparison results.

[0043] In one embodiment of the present invention, comparing candidate index values ​​under multiple preset dimensions corresponding to multiple candidate warehousing strategies pairwise and determining iterative warehousing strategies based on the comparison results includes: for any candidate warehousing strategies that are compared pairwise, if one candidate warehousing strategy is not inferior to another candidate warehousing strategy S2 under all preset dimensions, and one candidate warehousing strategy S1 is superior to another candidate warehousing strategy S2 under at least one preset dimension, then one candidate warehousing strategy S1 is added to the non-dominated strategy set; and the iterative warehousing strategy is determined based on the non-dominated strategy set.

[0044] For example, a candidate warehousing strategy S1 being no worse than another candidate warehousing strategy across all preset dimensions means that the positive metrics of candidate warehousing strategy S1 across any preset dimension are greater than or equal to those of another candidate warehousing strategy S2, and the negative metrics of candidate warehousing strategy S1 across any preset dimension are less than or equal to those of another candidate warehousing strategy S2. In other words, this holds true for all positive metrics. This exists for all contrarian indicators. This indicates that a candidate warehousing strategy S1 is not inferior to another candidate warehousing strategy S2 across all preset dimensions, where i is the i-th preset dimension, j is the j-th secondary indicator, S1 represents a candidate warehousing strategy, and S2 represents another warehousing strategy. Let S represent the j-th secondary indicator of a candidate warehousing strategy S1 under the i-th preset dimension.

[0045] For example, if a candidate warehousing strategy S1 is better than another candidate warehousing strategy S2 in at least one preset dimension, it means that the candidate index values ​​of all positive indicators of the candidate warehousing strategy S1 in at least one preset dimension are greater than the candidate index values ​​of the other candidate warehousing strategy S2, and the candidate index values ​​of all negative indicators are less than the candidate index values ​​of the other candidate warehousing strategy S2.

[0046] For example, if a candidate storage strategy is not inferior to another candidate storage strategy S2 in all preset dimensions, and a candidate storage strategy S1 is superior to another candidate storage strategy S2 in at least one preset dimension, then it means that candidate storage strategy S1 dominates candidate storage strategy S2, and candidate storage strategy S1 is added to the non-dominated strategy set. According to the above method, all candidate storage strategies are traversed, and all candidate storage strategies that are not dominated by any other strategy are selected and added to the non-dominated strategy set (if a candidate storage strategy S1 is found to be dominated by another candidate storage strategy Sn during the traversal, then a candidate storage strategy S1 is deleted from the non-dominated strategy set).

[0047] For example, candidate warehousing strategies from the set of non-dominated strategies are selected as the Pareto optimal set and sent to the user to demonstrate the best trade-off strategy between different optimization objectives. The user can select one or more strategies from the Pareto optimal set as iterative warehousing strategies according to actual business needs.

[0048] Step S250 repeats the iteration operation until the strategy score of the iterative storage strategy is greater than the preset score threshold, and the iterative storage strategy is taken as the optimized target storage strategy.

[0049] In one embodiment of the present invention, after repeatedly performing the iterative operation, the method further includes: if the number of times the iterative operation is repeatedly performed exceeds a preset threshold, then the repeated execution of the iterative operation is stopped, and the iterative storage strategy with the highest strategy score is determined as the target storage strategy.

[0050] For example, after multiple iterations, the large language model can be fine-tuned periodically to improve its performance in cause analysis and strategy generation in the warehousing domain. Simultaneously, the analysis and strategies generated by the large language model can be evaluated (e.g., scored and corrected), allowing for further optimization of the large language model based on the evaluation.

[0051] Please see Figure 3 , Figure 3 This is a block diagram of a warehouse strategy optimization device provided in one embodiment of the present invention. This device can be applied to... Figure 1 The implementation environment shown can also be applied to other exemplary implementation environments and specifically configured in other devices. This embodiment does not limit the implementation environment to which the device is applicable.

[0052] like Figure 3 As shown, the exemplary warehouse strategy optimization device includes: The initial strategy input module 310 is used to obtain the initial warehousing strategy, input the initial warehousing strategy into the preset warehousing simulation model, and obtain the initial index values ​​under multiple preset dimensions. The initial strategy scoring module 320 is used to obtain the strategy score of the initial warehousing strategy based on the initial indicator value and its corresponding preset weight value. The candidate strategy generation module 330 is used to input the initial indicator value, the initial storage strategy and the corresponding strategy score into the large language model to generate candidate storage strategies. The candidate strategy iteration module 340 is used to take the candidate storage strategy as the iterative storage strategy and perform iterative operations. The iterative operations include: inputting the iterative storage strategy into the storage simulation model to obtain the current indicator value under multiple preset dimensions; obtaining the strategy score of the iterative storage strategy based on the current indicator value and its corresponding preset weight value; if the strategy score of the iterative storage strategy is less than the preset score threshold, then inputting the current indicator value, the iterative storage strategy and the corresponding strategy score into the large language model again to generate a new candidate storage strategy, and using the new candidate storage strategy as the updated iterative storage strategy. The target strategy optimization module 350 is used to repeatedly perform iterative operations until the strategy score of the iterative storage strategy is greater than a preset score threshold, and then the iterative storage strategy is used as the optimized target storage strategy.

[0053] Through the aforementioned device, a more comprehensive and systematic strategy evaluation scheme, and the introduction of a large language model to generate diverse candidate strategies, more intelligent strategy optimization can be achieved, thereby improving the overall efficiency of the warehousing system.

[0054] It is understood that the warehousing strategy optimization device and the warehousing strategy optimization method provided in the above embodiments belong to the same concept. The specific execution method of the warehousing strategy optimization method has been described in detail in the above embodiments and will not be repeated here. In practical applications, the warehousing strategy optimization device provided in the above embodiments can be assigned to different functional modules as needed. That is, the internal structure of the warehousing strategy optimization device can be divided into different functional modules, and then all or part of the functions of the corresponding functional modules can be implemented by the warehousing strategy optimization method described in the above embodiments. No specific limitations are imposed here. For example, the initial strategy input module 310 includes steps S210 and related steps, the initial strategy scoring module 320 includes steps S220 and related steps, the candidate strategy generation module 330 includes steps S230 and related steps, the candidate strategy iteration module 340 includes steps S240 and related steps, and the target strategy optimization module 350 includes steps S250 and related steps.

[0055] Figure 4 This is a schematic diagram of an electronic device provided in one embodiment of the present invention. It should be noted that... Figure 4The computer system 400 of the electronic device shown is merely an example and should not impose any limitation on the functionality and scope of use of the embodiments of the present invention.

[0056] like Figure 4 As shown, the computer system 400 includes a Central Processing Unit (CPU) 401, which can perform various appropriate actions and processes based on programs stored in Read-Only Memory (ROM) 402 or programs loaded from storage portion 408 into Random Access Memory (RAM) 403, such as performing the methods described in the above embodiments. The RAM 403 also stores various programs and data required for system operation. The CPU 401, ROM 402, and RAM 403 are interconnected via a bus 404. An Input / Output (I / O) interface 405 is also connected to the bus 404.

[0057] The following components are connected to I / O interface 405: an input section 406 including a keyboard, mouse, etc.; an output section 407 including a cathode ray tube (CRT), liquid crystal display (LCD), etc., and speakers, etc.; a storage section 408 including a hard disk, etc.; and a communication section 409 including a network interface card such as a LAN (Local Area Network) card, modem, etc. The communication section 409 performs communication processing via a network such as the Internet. A drive 410 is also connected to I / O interface 405 as needed. A removable medium 411, such as a disk, optical disk, magneto-optical disk, semiconductor memory, etc., is installed on drive 410 as needed so that computer programs read from it can be installed into storage section 408 as needed.

[0058] In particular, according to embodiments of the present invention, the processes described above with reference to the flowcharts can be implemented as computer software programs. For example, embodiments of the present invention include a computer program product comprising a computer program carried on a computer-readable medium, the computer program containing a computer program for performing the methods shown in the flowcharts. In such embodiments, the computer program can be downloaded and installed from a network via communication section 409, and / or installed from removable medium 411. When the computer program is executed by central processing unit (CPU) 401, it performs various functions defined in the system of the present invention.

[0059] It should be noted that the computer-readable medium shown in the embodiments of the present invention can be a computer-readable signal medium or a computer-readable storage medium, or any combination thereof. A computer-readable storage medium can be, for example, an electrical, magnetic, optical, electromagnetic, infrared, or semiconductor system, apparatus, or device, or any combination thereof. More specific examples of a computer-readable storage medium may include, but are not limited to: an electrical connection having one or more wires, a portable computer disk, a hard disk, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM), flash memory, optical fiber, portable compact disc read-only memory (CD-ROM), optical storage device, magnetic storage device, or any suitable combination thereof. In the present invention, a computer-readable signal medium may include a data signal propagated in baseband or as part of a carrier wave, carrying a computer-readable computer program. Such propagated data signals may take various forms, including but not limited to electromagnetic signals, optical signals, or any suitable combination thereof. Computer-readable signal media can also be any computer-readable medium other than computer-readable storage media, which can send, propagate, or transmit a program for use by or in connection with an instruction execution system, apparatus, or device. The computer program contained on the computer-readable medium can be transmitted using any suitable medium, including but not limited to wireless, wired, etc., or any suitable combination thereof.

[0060] The flowcharts and block diagrams in the accompanying drawings illustrate the architecture, functionality, and operation of possible implementations of systems, methods, and computer program products according to various embodiments of the present invention. Each block in a flowchart or block diagram may represent a module, segment, or portion of code containing one or more executable instructions for implementing a specified logical function. It should also be noted that in some alternative implementations, the functions indicated in the blocks may occur in a different order than those indicated in the drawings. For example, two consecutively indicated blocks may actually be executed substantially in parallel, and they may sometimes be executed in reverse order. It should also be noted that each block in a block diagram or flowchart, and combinations of blocks in a block diagram or flowchart, may be implemented using a dedicated hardware-based system that performs the specified function or operation, or using a combination of dedicated hardware and computer instructions.

[0061] The units described in the embodiments of the present invention can be implemented in software or hardware, and the described units can also be located in a processor. The names of these units do not necessarily limit the specific unit itself.

[0062] Another aspect of the present invention provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a computer's processor, causes the computer to perform the warehouse strategy optimization method as described above. This computer-readable storage medium may be included in the electronic device described in the above embodiments, or it may exist independently and not incorporated into the electronic device.

[0063] Another aspect of the present invention provides a computer program product or computer program including computer instructions stored in a computer-readable storage medium. A processor of a computer device reads the computer instructions from the computer-readable storage medium and executes the computer instructions, causing the computer device to perform the warehousing strategy optimization method provided in the various embodiments described above.

[0064] The above embodiments are merely illustrative of the principles and effects of the present invention and are not intended to limit the invention. Any person skilled in the art can modify or alter the above embodiments without departing from the spirit and scope of the present invention. Therefore, all equivalent modifications or alterations made by those skilled in the art without departing from the spirit and technical concept disclosed in the present invention should still be covered by the claims of the present invention.

Claims

1. A method for optimizing warehousing strategies, characterized in that, The method includes: Obtain the initial warehousing strategy, input the initial warehousing strategy into the preset warehousing simulation model, and obtain the initial index values ​​under multiple preset dimensions; The strategy score of the initial warehousing strategy is obtained based on the initial indicator value and its corresponding preset weight value. The initial index value, the initial storage strategy, and the corresponding strategy score are input into the large language model to generate candidate storage strategies. The candidate warehousing strategy is used as an iterative warehousing strategy, and an iterative operation is performed. The iterative operation includes: inputting the iterative warehousing strategy into the warehousing simulation model to obtain the current indicator value under the multiple preset dimensions; obtaining the strategy score of the iterative warehousing strategy based on the current indicator value and its corresponding preset weight value; if the strategy score of the iterative warehousing strategy is less than a preset score threshold, then inputting the current indicator value, the iterative warehousing strategy and the corresponding strategy score into the large language model again to generate a new candidate warehousing strategy, and using the new candidate warehousing strategy as the updated iterative warehousing strategy. Repeat the iterative operation until the strategy score of the iterative storage strategy is greater than the preset score threshold, and then use the iterative storage strategy as the optimized target storage strategy.

2. The warehousing strategy optimization method according to claim 1, characterized in that, The initial warehousing strategy is input into a preset warehousing simulation model to obtain initial indicator values ​​under multiple preset dimensions, including: The process parameters and equipment configuration corresponding to the initial warehousing strategy are input into the warehousing simulation model to obtain initial index values ​​under multiple preset dimensions. The warehousing simulation model includes a digital twin model based on discrete event simulation. The process parameters include at least one of the following: characteristics of incoming materials, order patterns for outgoing goods, and storage strategy rules. The equipment configuration includes at least one of the following: warehouse layout structure, performance parameters of handling equipment, and attributes of storage equipment.

3. The warehousing strategy optimization method according to claim 2, characterized in that, The strategy score of the initial warehousing strategy, based on the initial indicator value and its corresponding preset weight value, includes: The initial index values ​​are normalized. The strategy score of the initial warehousing strategy is obtained by weighted summation of the normalized initial index values ​​and their corresponding preset weight values.

4. The warehousing strategy optimization method according to any one of claims 1-3, characterized in that, Before inputting the initial metric values, the initial warehousing strategy, and the corresponding strategy scores into the large language model, the following steps are also included: Obtain multi-source warehousing information, which includes at least historical warehousing operation data, professional documents in the warehousing field, and warehousing industry knowledge; The multi-source warehouse information is vectorized using text embedding technology to obtain text vectors. The text vectors are mapped and associated with entities in the preset knowledge graph of the large language model to enhance the knowledge of the large language model.

5. The warehousing strategy optimization method according to any one of claims 1-3, characterized in that, The initial metric value, the initial storage strategy, and the corresponding strategy score are input into the large language model to generate candidate storage strategies, including: Constructing prompt words that include optimization instructions, the prompt words also including the initial indicator value, the initial storage strategy, and the corresponding strategy score, the optimization instructions including instructions for guiding the large language model to optimize the initial indicator value under the multiple preset dimensions. The prompt words are input into the large language model to obtain at least one candidate storage strategy generated by the large language model based on the prompt words.

6. The warehousing strategy optimization method according to any one of claims 1-3, characterized in that, If there are multiple candidate warehousing strategies, then the candidate warehousing strategies used as iterative warehousing strategies include: Multiple candidate warehousing strategies are input into the warehousing simulation model to obtain candidate index values ​​under multiple preset dimensions corresponding to the multiple candidate warehousing strategies; The candidate index values ​​under multiple preset dimensions corresponding to the multiple candidate warehousing strategies are compared pairwise, and the iterative warehousing strategy is determined based on the comparison results.

7. The warehousing strategy optimization method according to any one of claims 1-3, characterized in that, After repeating the iterative operation, the process further includes: If the number of times the iterative operation is repeated exceeds a preset threshold, the iterative operation is stopped, and the iterative storage strategy with the highest strategy score is determined as the target storage strategy.

8. A warehouse strategy optimization device, characterized in that, The warehousing strategy optimization includes: The initial strategy input module is used to obtain the initial warehousing strategy, input the initial warehousing strategy into the preset warehousing simulation model, and obtain the initial index values ​​under multiple preset dimensions. The initial strategy scoring module is used to obtain the strategy score of the initial warehousing strategy based on the initial indicator value and its corresponding preset weight value. The candidate strategy generation module is used to input the initial indicator value, the initial storage strategy and the corresponding strategy score into the large language model to generate candidate storage strategies. The candidate strategy iteration module is used to take the candidate storage strategy as the iterative storage strategy and perform an iterative operation. The iterative operation includes: inputting the iterative storage strategy into the storage simulation model to obtain the current indicator value under the multiple preset dimensions; obtaining the strategy score of the iterative storage strategy based on the current indicator value and its corresponding preset weight value; if the strategy score of the iterative storage strategy is less than a preset score threshold, then inputting the current indicator value, the iterative storage strategy and the corresponding strategy score back into the large language model to generate a new candidate storage strategy, and using the new candidate storage strategy as the updated iterative storage strategy. The target strategy optimization module is used to repeatedly execute the iterative operation until the strategy score of the iterative storage strategy is greater than the preset score threshold, and then the iterative storage strategy is used as the optimized target storage strategy.

9. An electronic device, characterized in that, The electronic device includes: One or more processors; A storage device for storing one or more programs, which, when executed by one or more processors, cause the electronic device to implement the warehouse strategy optimization method as described in any one of claims 1-7.

10. A computer-readable storage medium, characterized in that, It stores a computer program that, when executed by the computer's processor, causes the computer to perform the warehouse strategy optimization method as described in any one of claims 1-7.

Images