Llm-based cross-border supply chain optimization method and device, electronic equipment and medium

Abstract

This application relates to the field of LLM-based cross-border supply chain optimization technology, disclosing an LLM-based method, apparatus, electronic device, and medium for cross-border supply chain optimization. The method includes: hierarchically parsing relevant legal texts to accurately extract legal rule features, mapping them to a preset mathematical space to form mathematical constraints, automatically identifying and optimizing initial suppliers that do not conform to the mathematical constraints, thereby optimizing the supply chain. The beneficial effects of this invention are: reduced legal risks, improved supply chain transparency and flexibility, cost savings and efficiency improvements, and the ability to automatically apply legal requirements to supply chain management, greatly improving the efficiency and accuracy of compliance audits.

Description

Technical Field

[0001] This invention relates to the field of LLM-based cross-border supply chain optimization technology, and in particular to an LLM-based cross-border supply chain optimization method, apparatus, electronic device, and medium. Background Technology

[0002] In the context of globalized trade, supply chain management faces increasingly complex legal and compliance requirements. As countries tighten regulations on issues such as product origin and trade laws, companies need to respond to legal challenges in real time. This not only affects operational efficiency but may also lead to legal disputes and economic losses. Traditional supply chain management methods are no longer adequate for the complexities of modern trade.

[0003] Currently, effective legal text analysis technology is lacking, and manual review is often relied upon, which is inefficient and prone to errors. In addition, existing supply chain management systems often fail to synchronize with the supply chain level when dealing with legal compliance issues, and also fail to effectively assess the compliance and risks of suppliers. Summary of the Invention

[0004] Therefore, it is necessary to propose an LLM-based method, apparatus, electronic device, and medium for optimizing existing cross-border supply chains.

[0005] A method for optimizing cross-border supply chains based on LLM, the method comprising:

[0006] Obtain the hierarchical supply chain and related legal documents for the specified product;

[0007] The legal text is hierarchically parsed using an LLM model to obtain legal rule features;

[0008] The legal rule features are mapped to a preset mathematical space to generate mathematical constraints in the preset mathematical space;

[0009] Obtain the initial suppliers at each level of the hierarchical supply chain, and obtain the corresponding supplier data according to the level of each initial supplier;

[0010] The supplier data is mapped into the preset mathematical space, and the supplier data is verified to meet the requirements of the mathematical constraints.

[0011] The initial suppliers whose verification results do not meet the requirements of the mathematical constraints are optimized so that the optimized hierarchical supply chain meets the requirements of the mathematical constraints.

[0012] Furthermore, the step of obtaining the hierarchical supply chain of the specified product includes:

[0013] The specified product is structurally decomposed to obtain the BOM tree structure of the specified product; wherein the BOM tree structure includes multiple nodes, and each node corresponds to a product component of the specified product;

[0014] Obtain the initial suppliers and importance of each node in the BOM tree structure;

[0015] A hierarchical supply chain for the specified product is generated based on the initial suppliers of each node and their corresponding importance.

[0016] Furthermore, after the step of structurally decomposing the specified product to obtain the BOM tree structure of the specified product, the method further includes:

[0017] Obtain the product components corresponding to each node;

[0018] Based on the product composition, obtain multiple alternative suppliers corresponding to the node;

[0019] Calculate the similarity score between the initial supplier and each of the alternative suppliers for each of the nodes;

[0020] The alternative suppliers are sorted according to the magnitude of their respective similarity scores.

[0021] Furthermore, after the step of obtaining multiple alternative suppliers corresponding to the node based on the product composition, the method further includes:

[0022] Receive the objective function and data information from each of the alternative suppliers;

[0023] The data and objective function are input into a preset mathematical space to optimize and solve the objective function, thereby obtaining the target supply chain.

[0024] By comparing the target supply chain with the hierarchical supply chain, comparison information is obtained;

[0025] The comparison information is sent to the designated terminal.

[0026] Furthermore, the step of performing hierarchical parsing of the legal text using an LLM model to obtain legal rule features includes:

[0027] The legal text is hierarchically parsed using a pre-defined semantic segmentation technique to divide it into multiple paragraph modules;

[0028] Obtain the category attributes of each paragraph module and extract the paragraph modules as clause paragraph modules;

[0029] A three-dimensional rule feature matrix is ​​constructed based on the aforementioned clause paragraph module to serve as the legal rule feature; wherein, the three-dimensional rule feature matrix includes a validity dimension, a timeliness dimension, and a geographical dimension.

[0030] Furthermore, after the steps of obtaining the initial suppliers at each level of the hierarchical supply chain and obtaining the corresponding supplier data according to the level of each initial supplier, the method further includes:

[0031] Obtain the specified information regarding the production location of the designated product and the location of the initial supplier;

[0032] The specified information is mapped to the preset mathematical space to generate the specified constraints of the preset mathematical space;

[0033] The supplier data is mapped into the preset mathematical space, and the supplier data is verified to meet the specified constraints.

[0034] The initial suppliers whose verification results show that they do not meet the requirements of the specified constraints are optimized so that the optimized initial suppliers meet the requirements of the specified constraints.

[0035] Furthermore, after the step of optimizing the initial suppliers whose verification results do not meet the requirements of the mathematical constraints so that the optimized hierarchical supply chain meets the requirements of the mathematical constraints, the method further includes:

[0036] Obtain the locations of each target supplier in the optimized hierarchical supply chain, as well as the production location of the specified product;

[0037] Obtain the corresponding smart compliance certificate template based on the production location and each of the aforementioned locations;

[0038] Obtain the compliance data corresponding to each of the smart compliance certificate templates and fill it into the smart compliance certificate templates to obtain the smart compliance certificates corresponding to each target supplier.

[0039] A cross-border supply chain optimization device based on LLM, the device comprising:

[0040] The legal document acquisition module is used to acquire the hierarchical supply chain and related legal documents for a specified product.

[0041] The legal rule feature acquisition module is used to perform hierarchical parsing of the legal text using an LLM model to obtain legal rule features;

[0042] A mathematical constraint generation module is used to map the legal rule features to a preset mathematical space to generate mathematical constraints in the preset mathematical space.

[0043] The supplier data acquisition module is used to acquire the initial suppliers at each level of the hierarchical supply chain, and to acquire the corresponding supplier data according to the level of each initial supplier.

[0044] The supplier data mapping module is used to map supplier data into the preset mathematical space and verify that the supplier data meets the requirements of the mathematical constraints.

[0045] The first optimization module is used to optimize the initial suppliers whose verification results do not meet the requirements of the mathematical constraints, so that the optimized hierarchical supply chain meets the requirements of the mathematical constraints.

[0046] An electronic device includes a memory and a processor, the memory storing a computer program that, when executed by the processor, causes the processor to perform the following steps:

[0047] Obtain the hierarchical supply chain and related legal documents for the specified product;

[0048] The legal text is hierarchically parsed using an LLM model to obtain legal rule features;

[0049] The legal rule features are mapped to a preset mathematical space to generate mathematical constraints in the preset mathematical space;

[0050] Obtain the initial suppliers at each level of the hierarchical supply chain, and obtain the corresponding supplier data according to the level of each initial supplier;

[0051] The supplier data is mapped into the preset mathematical space, and the supplier data is verified to meet the requirements of the mathematical constraints.

[0052] The initial suppliers whose verification results do not meet the requirements of the mathematical constraints are optimized so that the optimized hierarchical supply chain meets the requirements of the mathematical constraints.

[0053] A computer-readable storage medium storing a computer program, which, when executed by a processor, causes the processor to perform the following steps:

[0054] Obtain the hierarchical supply chain and related legal documents for the specified product;

[0055] The legal text is hierarchically parsed using an LLM model to obtain legal rule features;

[0056] The legal rule features are mapped to a preset mathematical space to generate mathematical constraints in the preset mathematical space;

[0057] Obtain the initial suppliers at each level of the hierarchical supply chain, and obtain the corresponding supplier data according to the level of each initial supplier;

[0058] The supplier data is mapped into the preset mathematical space, and the supplier data is verified to meet the requirements of the mathematical constraints.

[0059] The initial suppliers whose verification results do not meet the requirements of the mathematical constraints are optimized so that the optimized hierarchical supply chain meets the requirements of the mathematical constraints.

[0060] The beneficial effects of this invention are as follows: By hierarchically parsing relevant legal texts, the features of legal rules are accurately extracted and then mapped onto a preset mathematical space to form mathematical constraints. This automatically identifies and optimizes initial suppliers that do not meet these constraints, ensuring the compliance and stability of the entire supply chain. This reduces legal risks, enhances the transparency and flexibility of the supply chain, achieves cost savings and efficiency improvements, and enables the automated application of legal requirements to supply chain management, greatly improving the efficiency and accuracy of compliance audits. Attached Figure Description

[0061] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0062] in:

[0063] Figure 1 This is a diagram illustrating the application environment of an LLM-based cross-border supply chain optimization method in one embodiment.

[0064] Figure 2 This is a flowchart of a cross-border supply chain optimization method based on LLM in one embodiment;

[0065] Figure 3 This is a structural block diagram of a cross-border supply chain optimization device based on LLM in one embodiment;

[0066] Figure 4 This is a structural block diagram of an electronic device in one embodiment. Detailed Implementation

[0067] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0068] Figure 1 This is a diagram illustrating an LLM-based cross-border supply chain optimization application environment in one embodiment. (Refer to...) Figure 1 This LLM-based cross-border supply chain optimization method is applied to an LLM-based cross-border supply chain optimization system. The system includes a terminal 110 and a server 120. The terminal 110 and server 120 are connected via a network. The terminal 110 can be a desktop terminal or a mobile terminal, specifically a mobile phone, tablet, laptop, or other similar device. The server 120 can be a standalone server or a server cluster consisting of multiple servers. The terminal 110 is used to obtain the hierarchical supply chain for a specified product, and the server 120 is used to optimize the hierarchical supply chain.

[0069] like Figure 2 As shown, in one embodiment, an LLM-based cross-border supply chain optimization method is provided. This method can be applied to both terminals and servers; this embodiment illustrates its application to terminals. The LLM-based cross-border supply chain optimization method specifically includes the following steps:

[0070] S1: Obtain the hierarchical supply chain and related legal documents for the specified product;

[0071] S2: The legal text is hierarchically parsed using an LLM model to obtain legal rule features;

[0072] S3: Map the legal rule features to a preset mathematical space to generate mathematical constraints in the preset mathematical space;

[0073] S4: Obtain the initial suppliers at each level in the hierarchical supply chain, and obtain the corresponding supplier data according to the level of each initial supplier;

[0074] S5: Map the supplier data to the preset mathematical space and verify that the supplier data meets the requirements of the mathematical constraints;

[0075] S6: Optimize the initial suppliers whose verification results do not meet the requirements of the mathematical constraints so that the optimized hierarchical supply chain meets the requirements of the mathematical constraints.

[0076] This LLM-based cross-border supply chain optimization method can be widely applied to the manufacturing and distribution industries involved in cross-border exports. In manufacturing, origin verification faces three major pain points: supply chain penetration challenges (difficulty in tracing materials from multi-tiered suppliers, such as second-tier chip suppliers in a mobile phone assembly plant in region A); dynamic rule adaptation (difficulty in responding in real-time to updates to the "cumulative rules" in agreements); and composite standard determination (lack of intelligent decision support when simultaneously meeting both cost classification changes and regional value composition standards), leading to complex supply chain management and legal compliance risks for enterprises. This method extracts legal rule features through hierarchical analysis of legal texts and maps them to a mathematical space, helping enterprises formulate compliant supply chain strategies. For example, for a specific product, an enterprise can identify and optimize initial suppliers that do not comply with legal constraints, thereby reducing compliance risks and improving the flexibility and reliability of the supply chain. In the distribution industry, an optimized supply chain ensures timely delivery of goods while complying with industry regulations, improving customer satisfaction.

[0077] As described in step S1 above, the hierarchical supply chain and related legal documents for the specified product are obtained. The product requiring optimization is identified, and hierarchical supply chain information related to that product is extracted through connections to the enterprise's internal ERP (Enterprise Resource Planning) system or other databases. A hierarchical supply chain typically includes all relevant nodes from raw material suppliers to the final product assembly plant, namely Tier 1 (direct suppliers), Tier 2, Tier 3, and further downstream suppliers. This structured information acquisition is a fundamental element of supply chain management, helping to identify the function and responsibility of each node in the hierarchical supply chain. Furthermore, the system collects all legal documents related to the product, including trade agreements, regional laws and regulations, industry standards, and international treaties. By centrally acquiring these legal documents, the system lays the data foundation for subsequent analysis and application.

[0078] As described in step S2 above, the legal text is hierarchically parsed using an LLM model to obtain legal rule features. Natural Language Processing (NLP) technology is used to perform hierarchical parsing of the acquired legal text. Specifically, methods such as semantic segmentation and dependency parsing are employed to progressively break the legal text down into modules such as clause sections, annotation sections, and example sections. Through this hierarchical parsing, the system can extract key information from the legal clauses, such as mandatory clauses, referential clauses, and their accompanying exceptions. For example, the system can identify core constraints related to certificates of origin and regional value components. After the legal text parsing is completed, the system will generate a document containing all legal rule features, which will provide a practical basis for subsequent mapping to mathematical space and further optimization decisions. Specifically, hierarchical parsing of legal text using a Large Language Model (LLM) can effectively extract and understand the hidden legal rule features in the text. Legal texts are typically characterized by complex structures, specialized terminology, and high abstraction, making traditional manual parsing methods inefficient and prone to errors. LLM, with its powerful natural language processing capabilities, can fully leverage contextual information and a rich knowledge base to conduct in-depth analysis of legal texts. The hierarchical parsing process typically includes sentence segmentation, paragraphing, and semantic understanding of the legal text. During this process, LLM identifies key information in the text, such as legal clauses, scope of application, obligations, and rights. This extracted information can be further summarized into a hierarchical structure, such as legal rules, sub-rules, and their applicable conditions. This hierarchical structure helps to clearly demonstrate the relationships between legal rules. By acquiring legal rule features, LLM can support more advanced legal research and compliance analysis. For example, when formulating compliance policies and supply chain strategies, companies can use the extracted information to ensure compliance with relevant laws and regulations and reduce legal risks. Furthermore, the extracted hierarchical features can provide important evidence for subsequent automatic generation of legal documents, judgment analysis, and related decision-making, promoting the intelligent and automated process of legal services. Therefore, LLM's hierarchical parsing not only improves the processing efficiency of legal texts but also provides strong support for intelligent decision-making in legal practice.

[0079] As described in step S3 above, the legal rule features are mapped to a preset mathematical space to generate mathematical constraints for that space. By mapping the extracted legal rule features to the preset mathematical space, effective mathematical constraints are generated, defining quantitative standards of different dimensions. The system can describe various constraints in legal provisions. For example, indicators such as regional value components and product localization rates can be expressed through mathematical formulas, forming a multi-dimensional mathematical constraint model. This model not only provides enterprises with compliance inspection standards but also allows for rapid adjustments and optimizations when laws and regulations are updated in real time, ensuring that the supply chain always complies with current regulatory requirements. Furthermore, this mapping process can also utilize more complex mathematical tools such as tensor analysis, effectively handling and optimizing multiple constraints, thereby providing enterprises with more accurate compliance assessments. The preset mathematical space is specifically a multi-dimensional space based on vector or tensor structures, used to represent the quantitative constraints of legal rules.

[0080] As described in step S4 above, the initial suppliers at each level of the hierarchical supply chain are obtained, and corresponding supplier data is acquired based on the level of each initial supplier. This requires delving into the hierarchical supply chain to identify and classify each supplier. Based on the acquired hierarchical supply chain information, the system sequentially accesses the databases related to each supplier to extract their basic information and relevant data. This information includes the supplier's geographical location, main supplied products, price, delivery cycle, compliance records, etc. By acquiring initial supplier data layer by layer, the system can conduct a more detailed evaluation of each node. For example, the violations of Tier 1 suppliers may directly affect the compliance of the finished product, thus requiring a deeper compliance audit. The information of Tier 2 and Tier 3 suppliers is equally important; their compliance status and relative cost-effectiveness will affect the optimization effect of the hierarchical supply chain.

[0081] As described in step S5 above, supplier data is mapped to the preset mathematical space, and the supplier data is verified to meet the requirements of the mathematical constraints. The obtained supplier data is mapped to the preset mathematical space and compared and verified with the corresponding mathematical constraints. This verification process is used to accurately assess the compliance improvement of each supplier. Specifically, all collected supplier data is substituted into the established mathematical model to calculate its compliance status in terms of legal compliance, such as calculating the supplier's regional value component and compliance effectiveness. During this process, if a supplier's data is found to be inconsistent with the preset mathematical constraints, the system will automatically mark it as a non-compliant supplier and generate a detailed compliance report for it. This function not only quickly identifies potential risky suppliers but also provides accurate data for subsequent optimization. Simultaneously, this process ensures the transparency of the hierarchical supply chain, enabling enterprises to monitor the compliance of each link in real time, adjust business strategies promptly, and minimize legal risks.

[0082] As described in step S6 above, initial suppliers whose verification results do not meet the requirements of the mathematical constraints are optimized to ensure that the optimized hierarchical supply chain complies with the mathematical constraints. Optimization of initial suppliers whose verification results do not meet the mathematical constraints ensures that the final hierarchical supply chain complies with all legal compliance requirements. The optimization process can take various forms, such as replacing non-compliant suppliers with new compliant supply chain members, or providing guidance and support to existing suppliers to improve their compliance. Alternatively, the auditing of these non-compliant suppliers can be strengthened, feedback collected, and targeted improvement measures implemented. After optimization, the system reassesses the entire hierarchical supply chain structure to ensure that all nodes participating in the hierarchical supply chain comply with the latest mathematical constraints. This optimization step not only helps reduce legal risks but also enhances the company's competitiveness by reducing overall procurement costs and improving product quality. Simultaneously, the optimization results and their impact are incorporated into the system's feedback mechanism for continuous improvement and optimization in future supply chain management, forming an efficient and compliant supply chain ecosystem.

[0083] In one embodiment, step S1 of obtaining the hierarchical supply chain of a specified product includes:

[0084] S101: Perform structural decomposition on the specified product to obtain the BOM tree structure of the specified product; wherein, the BOM tree structure includes multiple nodes, and each node corresponds to a product component of the specified product;

[0085] S102: Obtain the initial suppliers of each node in the BOM tree structure and the importance of each node;

[0086] S103: Generate a hierarchical supply chain for the specified product based on the initial suppliers of each node and their corresponding importance.

[0087] As described in step S101 above, the specified product is structurally decomposed to obtain its BOM tree structure. The BOM tree structure is a graphical representation of the product's components and their hierarchical relationships, typically displayed in a tree structure. Each node represents a component or part of the product. For example, a smartphone's BOM tree might include multiple levels such as display, processor, battery, and casing. During this process, the system integrates with the enterprise's ERP system or other relevant databases to extract all components of the product and identify the physical and logical relationships between each component. Generating this tree structure helps to clearly demonstrate the product's hierarchical relationships and clarify the position and function of each component within the overall product. Simultaneously, the construction of the BOM tree provides foundational data for subsequent hierarchical supply chain analysis, enabling the system to more easily track each component and assess the impact of each node in the hierarchical supply chain on the final product's quality and compliance.

[0088] As described in step S102 above, the initial suppliers and importance of each node in the BOM tree structure are obtained. An initial supplier refers to a company responsible for providing a specific component or component. For example, the display screen of a smartphone might be supplied by a liquid crystal display (LCD) manufacturer; this company is the initial supplier of the display screen node. The importance assessment of a node is based on multiple factors, such as the impact of the component or component on the final product's performance, cost, and compliance. A common assessment principle is that critical components have relatively high importance; therefore, suppliers of core components such as processors and batteries are marked as high-priority nodes. Packaging materials or other auxiliary components, because they have a smaller impact on the product, can be considered less important. Through this importance assessment method, the system can ultimately identify which nodes occupy a key position in the hierarchical supply chain relationship, better supporting subsequent hierarchical supply chain optimization measures.

[0089] As described in step S103 above, based on the importance information of the initial suppliers and nodes in the BOM tree structure obtained in the previous steps, a complete hierarchical supply chain is generated. This process involves integrating all initial suppliers and their relationships with various nodes of the product to form a hierarchical supply chain model. In the hierarchical supply chain, Tier 1 suppliers directly provide parts or components to the product manufacturer, while Tier 2, Tier 3, and other suppliers provide secondary parts or raw materials. The generated hierarchical supply chain not only shows the position of different suppliers in the hierarchical supply chain but also clarifies their functional and responsibility relationships with the final product. In this way, the system can clearly express the supply relationships and dependencies between each level, providing the system with the objectives for evaluating and adjusting each link in the hierarchical supply chain. Ultimately, once the overall hierarchical supply chain is formed, enterprises can make adjustments according to actual conditions to ensure the scientific nature of supplier selection and management, effectively reduce delivery risks and compliance costs, and improve overall production efficiency and market competitiveness.

[0090] In one embodiment, after step S101 of structural decomposition of the specified product to obtain the BOM tree structure of the specified product, the method further includes:

[0091] S1021: Obtain the product components corresponding to each node;

[0092] S1022: Obtain multiple alternative suppliers corresponding to the node based on the product composition;

[0093] S1023: Calculate the similarity score between the initial supplier and each of the alternative suppliers for each of the nodes;

[0094] S1024: Sort the alternative suppliers according to the size of the similarity scores.

[0095] As described in step S1021 above, the product components corresponding to each node are obtained. The system extracts the product components corresponding to each node from the previously generated BOM tree structure. Whether it is an electronic product, mechanical equipment, or other types of product, each component represents the specific materials or components required. The information of these product components is very important because it directly relates to the product's performance, reliability, and compliance. By analyzing the BOM tree, the system can obtain the specific materials required for each node and combine them with the initial supplier information to further evaluate the effectiveness of the hierarchical supply chain. For example, the motherboard components of a mobile phone may include product components such as CPU, memory, and wireless module. This process usually involves the integration of multiple data sources, including ERP (Enterprise Resource Planning) systems, product design documents, and supplier documents. After extracting the product components, the system can better understand which components are key components and which are secondary components. This not only lays the foundation for subsequent alternative supplier acquisition and similarity assessment but also ensures the scientific and effective nature of the entire hierarchical supply chain in the material selection and supplier management process.

[0096] As described in step S1022 above, based on the extracted product components for each node, multiple alternative suppliers are identified and acquired. Alternative suppliers are those capable of providing similar or identical product components, possessing comparable technology and supply capabilities to the initial supplier. The introduction of alternative suppliers primarily aims to enhance the flexibility and resilience of the hierarchical supply chain and reduce risks associated with relying on a single supplier. By querying various information sources such as industry databases, market research, and historical procurement data, the system can effectively identify multiple alternative suppliers matching each product component. In most cases, this process also needs to consider factors such as the alternative supplier's geographical location, quality certification, delivery capacity, and price to ensure that the selected alternative supplier can meet actual production needs. Through this alternative supplier acquisition mechanism, enterprises can quickly adjust to unforeseen circumstances such as supplier disruptions, product quality issues, or price fluctuations, reducing the impact on overall production and compliance. Alternative suppliers can be obtained by "querying supplier databases or external market data platforms."

[0097] As described in step S1023 above, the similarity score between the initial supplier and each of the alternative suppliers for each node is calculated. The system needs to calculate the similarity between the initial supplier and its corresponding alternative supplier for each node. This process is a crucial step in quantitatively evaluating supplier capabilities and adaptability, aiming to provide data support for subsequent ranking and decision-making. The calculation of the similarity score involves multiple dimensions, including the supplier's historical performance, product quality, delivery speed, and price competitiveness. For example, a weighted scoring model can be used to quantitatively evaluate these factors and obtain a comprehensive similarity score. Specifically, the system first sets the weight of each evaluation indicator, then scores each supplier in detail to obtain the similarity score calculation formula.

[0098] As described in step S1024 above, alternative suppliers are ranked according to the calculated similarity scores. The purpose of this ranking is to determine which alternative suppliers have the greatest potential to become the initial supplier, thereby supporting decision-making. The ranking results will make it easier for managers to identify the most suitable alternative suppliers and make quick adjustments based on actual circumstances. Specific implementation methods typically involve inputting all alternative suppliers and their similarity scores into a ranking algorithm, which can be a simple quicksort or mergesort, etc. The system first checks the similarity scores of all suppliers and then outputs a ranked supplier list in descending order. In the final ranking, the supplier with the highest score will be given priority consideration as an alternative or replacement for the initial supplier. This process not only enhances the flexibility of the hierarchical supply chain but also ensures timely response to various changes in the hierarchical supply chain in actual operation, such as legal compliance risks, unstable market dynamics, or sudden supply disruptions, enabling the company to maintain competitiveness in a highly competitive market environment.

[0099] In one embodiment, after step S1022 of obtaining multiple alternative suppliers corresponding to the node based on the product composition, the method further includes:

[0100] S10231: Receive the objective function and data information of each of the alternative suppliers;

[0101] S10232: Input each of the data information and the objective function into a preset mathematical space to optimize and solve the objective function to obtain the target supply chain;

[0102] S10233: Compare the target supply chain with the hierarchical supply chain to obtain comparison information;

[0103] S10234: Send the comparison information to the designated terminal.

[0104] As described in step S10231 above, the system receives the objective function and data information from each of the alternative suppliers. The objective function is a mathematical expression used to measure system performance during optimization. It typically involves indicators such as cost, delivery time, compliance, or risk level of the target supply chain. For example, a company might want to minimize total procurement costs while maximizing product quality stability. For effective optimization, the system needs to define the objective function and ensure it aligns with business needs. The received alternative supplier data typically includes relevant characteristics of each supplier, such as delivery time, price, historical performance, compliance, and reputation. This data will be used to analyze the impact of different alternative suppliers on the objective function, providing a basis for subsequent optimization. In most cases, this data is stored in a structured format in a database, allowing for quick retrieval and processing. By integrating the objective function and alternative supplier data, the system lays a solid foundation for further optimization steps, ensuring the discovery of the optimal supply chain solution. The objective function is a function set by relevant personnel, such as minimizing total cost.

[0105] As described in step S10232 above, the received objective function and data information of alternative suppliers are input into a preset mathematical space for optimization. This process typically involves using mathematical modeling techniques, such as linear programming, integer programming, or other suitable algorithms, to optimize cost, efficiency, or other performance indicators. In this way, the system can generate an ideal target supply chain, i.e., a supply chain structure with optimal performance under specific constraints. Specifically, the system first transforms the objective function into a standard mathematical form and inputs data information of alternative suppliers, which may include suppliers' delivery capabilities, prices, compliance records, etc. Subsequently, the system performs calculations using mathematical algorithms to find the optimal combination of the target supply chain while meeting all business conditions and legal compliance requirements. The result of the optimization solution will be a specific supply chain configuration scheme, including the selected alternative suppliers and related ordering strategies.

[0106] As described in step S10233 above, the generated target supply chain is compared with the original hierarchical supply chain to obtain comparative information. Relevant parameters of the target and original hierarchical supply chains are extracted, including indicators such as cost, delivery time, compliance, and supplier risk, and compared one by one. Through this comparison, the system can identify the strengths and weaknesses of the target supply chain and understand the extent to which the optimization plan can improve the overall supply chain efficiency and compliance. For example, if the total cost of the target supply chain is significantly lower than that of the hierarchical supply chain, or if the delivery speed is significantly improved, the system will record these comparison results. In some cases, if the compliance of the target supply chain declines, the system will also record this negative information. The comparison results will form a detailed comparison report for decision-makers to refer to.

[0107] As described in step S10234 above, the obtained comparative information is sent to a designated terminal for relevant personnel to view and make decisions. This terminal can be a management computer, mobile device, or a terminal interface in an integrated information management system. In this way, key decision-makers can quickly obtain evaluation results and recommendations regarding the target supply chain and make more accurate strategic judgments. The sent comparative information typically includes specific comparative data and charts between the target supply chain and the hierarchical supply chain in terms of cost, efficiency, compliance, and other aspects. This intuitive information display helps decision-makers quickly understand the advantages and disadvantages of different solutions and select the best solution to implement.

[0108] In one embodiment, step S2, which involves hierarchically parsing the legal text using an LLM model to obtain legal rule features, includes:

[0109] S201: The legal text is hierarchically parsed using a preset semantic segmentation technology to divide the legal text into multiple paragraph modules;

[0110] S202: Obtain the category attribute of each paragraph module and extract the paragraph module as a clause paragraph module;

[0111] S203: Construct a three-dimensional rule feature matrix based on the clause paragraph module to serve as the legal rule feature; wherein, the three-dimensional rule feature matrix includes an effectiveness dimension, a timeliness dimension, and a geographical dimension.

[0112] As described in step S201 above, the legal text is hierarchically parsed using a pre-defined semantic segmentation technique. This process aims to effectively divide the legal text into multiple easily manageable and analyzable paragraph modules. Legal texts are often relatively complex, containing multi-layered information. Directly analyzing the entire text may lead to inefficiency and misunderstanding. Therefore, semantic segmentation technology can divide legal documents into paragraphs with independent meaning, such as clauses, definitions, exceptions, and annotations, based on context and semantic application strategies. This technique is typically based on Natural Language Processing (NLP) algorithms, which can identify logical relationships between sentences and extract important information. In this way, different legal clauses and conditions in the text are effectively extracted, forming multiple independent and clearly structured paragraph modules. This not only helps with subsequent rule extraction and analysis but also facilitates the citation and retrieval of specific legal clauses in subsequent work. For example, a clause in regulations regarding subsidies may be treated as a paragraph module with its contextual information intact. This approach improves the efficiency of legal text processing while reducing errors that may occur during manual processing, ensuring that accurate and reliable legal rule features are ultimately obtained.

[0113] As described in step S202 above, category attributes are obtained for each segmented paragraph module, and clause paragraph modules are further extracted. By analyzing the content and semantics of each paragraph, they are classified into specific legal categories, such as defining clauses, mandatory clauses, referential clauses, and exception clauses. Obtaining category attributes involves content analysis of each paragraph module, typically using text classification methods, such as machine learning-based classifiers or simple classification methods based on keywords and context rules. Determining whether a paragraph belongs to a specific type of legal clause may involve comparison with previous cases, semantic analysis, and inference of contextual relationships. Simultaneously, extracting clause paragraph modules from paragraphs helps the system understand the specific meaning and applicable scenarios of legal rules. Through this classification, the system not only improves its ability to parse legal texts but also provides strong data support for optimizing supply chain compliance.

[0114] As described in step S203 above, the extracted clause paragraph modules are used to construct a three-dimensional rule feature matrix. This matrix serves as a concrete representation of the acquired legal rule features, aiming to effectively organize and quantify complex legal clause information. The three-dimensional rule feature matrix typically includes three dimensions: validity, timeliness, and geographic scope. Each dimension categorizes and quantifies the attributes of legal clauses, reflecting the characteristics and applicability of legal rules in different aspects. Validity dimension: reflects the mandatory or referential nature of the clause. For example, some clauses in an ordinance are mandatory, while others are advisory, and some clauses may have exceptions. Timeliness dimension: shows the effective and applicable time of different clauses. For example, some legal clauses are only valid within a specific year, or there may be transitional provisions before a new revised clause takes effect. Geographic scope dimension: indicates the geographical scope of application of the legal clause, which may involve unique requirements of different countries, regions, or economic zones. For example, a clause may only be valid in a specific country or region. The validity dimension characterizes the strength and binding degree of the rule, and its value and encoding are as follows: Value: Mandatory: This rule must be followed and cannot be violated. Recommended: This rule is a suggestion; compliance is optional and depends on the specific circumstances. Exceptional: This rule applies to specific situations or individuals and is generally not universally applicable. Code: Mandatory = 1; Recommended = 2; Exceptional = 3; The time-limited dimension represents the applicable time limit of the rule, with the following values ​​and codes: Values: Perpetual: The rule is valid indefinitely within its validity period. Temporary: The rule is valid within a specific time frame, usually limited to start and end dates. Expired: The rule has expired and is no longer applicable. Codes: Perpetual = 1; Temporary = 2; Expired = 3. The geographical dimension represents the geographical scope of the rule's application, with the following values ​​and codes: Values: National: This rule applies to all regions globally. Regional: This rule applies to a specific region or province / city. Local: This rule applies only to a specific city or organization. Codes: Global = 1; Regional = 2; Local = 3.

[0115] In this way, the three-dimensional rule feature matrix constructed by the system not only organizes legal provisions in terms of data structure, but also provides foundational data for subsequent automated compliance checks and decision-making. Using this matrix, the system can process information from different dimensions in parallel, perform intelligent analysis and reasoning, and achieve efficient application of legal rules in the hierarchical supply chain optimization process.

[0116] In one embodiment, after step S4 of obtaining the initial suppliers at each level of the hierarchical supply chain and obtaining the corresponding supplier data according to the level of each initial supplier, the method further includes:

[0117] S501: Obtain the specified information regarding the production location of the designated product and the location of the initial supplier;

[0118] S502: Map the specified information to the preset mathematical space to generate the specified constraints of the preset mathematical space;

[0119] S503: Map the supplier data to the preset mathematical space and verify that the supplier data meets the requirements of the specified constraints;

[0120] S504: Optimize the initial supplier whose verification result does not meet the requirements of the specified constraints so that the optimized initial supplier meets the requirements of the specified constraints.

[0121] As described in step S501 above, the system collects and integrates regulatory information related to the production location and the locations of each initial supplier for the specified product. This information is crucial for understanding the compliance risks, trade barriers, and legal environments of different geographical regions. Regulatory factors directly impact the stability and reliability of the tiered supply chain, and the system needs to assess which regions' risks might affect product delivery and compliance. In practice, the system can obtain this information through multiple channels, including regulatory information analysis databases, relevant reports published by designated agencies, industry analysis data, and market research from professional consulting firms. The system will collect information such as whether there are compliance risks in the countries where the initial suppliers are located. This information helps to form a comprehensive risk assessment for each supplier, supporting subsequent optimization decisions and ensuring that risk factors affecting major supply decisions are fully considered during the optimization of the tiered supply chain, thereby improving the overall security and compliance of the tiered supply chain.

[0122] As described in step S502 above, the collected regulatory information is mapped to a pre-defined mathematical space to generate regulatory constraints. This process involves quantifying and standardizing the risks of the examined rule information, enabling effective management and analysis within a given mathematical framework. Typically, the mapping of regulatory information creates a set of variables, which may include geopolitical risk scores, compliance with trade rules, and diplomatic stability scores. To achieve this mapping, data modeling methods, such as multiple linear regression analysis or support vector machines (SVM), can be used to transform the rule information risks into a form suitable for processing in a mathematical space. Specifically, a regional risk scoring model can be used to score the regulatory information, which is then transformed into a mathematical form. For example, a risk scoring system can be set up, using weighted combinations of various information to form a comprehensive score, and these scores will correspond to specific constraints in the mathematical space. In this way, the system can identify which regions where initial suppliers are located have higher compliance risks in the current context and integrate this information as regulatory constraints into subsequent hierarchical supply chain management and optimization strategies. This provides stronger data guidance for hierarchical supply chain decisions, ensuring that companies consider relevant policies and risk factors when selecting partners, thereby protecting the interests of companies in the international trade environment.

[0123] As described in step S503 above, supplier data is mapped to the preset mathematical space, and the supplier data is verified to meet the requirements of the prescribed constraints. The collected supplier data is mapped to the preset mathematical space to verify whether it conforms to the previously set prescribed constraints. In specific operation, the system first extracts relevant supplier data, such as the supplier's geographical location, trade history, and the stability of past cooperation. Then, the system uses the established prescribed constraints to verify this data in real time. This process may include quantitative analysis and logical judgment to determine whether the regulatory risks in the supplier's region will affect its delivery capabilities. For example, if a supplier is located in a high-risk area, the system will mark it and warn the user that the supplier has a high geographical risk. Through this verification, the system can not only effectively screen compliant suppliers but also make timely adjustments to ensure the health of the overall hierarchical supply chain at the legal and policy levels.

[0124] As described in step S504 above, initial suppliers whose verification results show non-compliance with regulatory constraints are optimized. This process includes assessing the characteristics of non-compliant suppliers and identifying specific strategies for replacement, modification, or improvement to ensure a compliant supply chain hierarchy that ultimately meets regulatory constraints. Specific optimization methods may include diverse strategies, such as finding new alternative suppliers, modifying supplier management processes, or communicating with existing suppliers to improve their compliance. Simultaneously, the system can utilize historical data analysis techniques to assess the background, stability, and past delivery capabilities of potential alternative suppliers, thereby selecting the most suitable replacement. Furthermore, by establishing closer partnerships with compliant suppliers, companies can not only mitigate risks but also improve the flexibility and responsiveness of their supply chain hierarchy.

[0125] In one embodiment, after step S6, which optimizes the initial supplier whose verification result does not meet the requirements of the mathematical constraint so that the optimized hierarchical supply chain meets the requirements of the mathematical constraint, the method further includes:

[0126] S701: Obtain the location of each target supplier in the optimized hierarchical supply chain, as well as the production location of the specified product;

[0127] S702: Obtain the corresponding smart compliance certificate template based on the production location and each of the aforementioned locations;

[0128] S703: Obtain the compliance data corresponding to each of the smart compliance certificate templates and fill it into the smart compliance certificate template to obtain the smart compliance certificate corresponding to each target supplier.

[0129] As described in step S701 above, the locations of each target supplier in the optimized hierarchical supply chain, as well as the production location of the specified product, are obtained. Obtaining the geographical locations of all target suppliers in the optimized hierarchical supply chain and determining the production location of the specified product involves extracting relevant data from the enterprise's ERP system or supply chain management system. This data typically includes the target supplier's detailed address, country, region, etc. On the other hand, the production location information also needs to be extracted from the corresponding product data. For example, regulations and trade barriers in different countries or regions may affect manufacturers' compliance requirements. By compiling and listing this geographical information, the system can provide necessary background support for the subsequent generation of compliance certification templates, laying the foundation for ensuring the compliance of each target supplier.

[0130] As described in step S702 above, based on the previously obtained production location and target supplier location, the system retrieves the corresponding intelligent compliance certificate templates. Retrieving compliance certificate templates typically requires combining information from legal databases, industry standards, and designated agencies. The system will quickly identify the most suitable compliance certificate template based on the information about the production location and supplier location. For example, if a product is manufactured in region A, and the target supplier is located in region B, the system can reference the corresponding compliance certificate template under the free trade agreement between region A and region B. These templates usually contain standard legal text and structure to effectively present the required compliance information.

[0131] As described in step S703 above, the system acquires compliance data corresponding to each smart compliance certificate template and fills this data into the acquired templates to generate smart compliance certificates for each target supplier. Specifically, the system first collects applicable compliance data from the company database, financial records, audit reports, etc., which may include information such as product certificates of origin, material compliance, production standards, and supply chain transparency. This data will be filled into the corresponding compliance certificate templates by the system to generate compliant certificate documents. When processing the data, the system must ensure the accuracy and consistency of the information entered to avoid legal and compliance risks due to data errors. Relevant data is automatically extracted from the compliance database and populated into the templates using interfaces or web scraping technology.

[0132] Reference Figure 3 The present invention also provides an LLM-based cross-border supply chain optimization device, the device comprising:

[0133] The legal text acquisition module 902 is used to acquire the hierarchical supply chain and related legal texts of a specified product.

[0134] The legal rule feature acquisition module 904 is used to perform hierarchical parsing of the legal text using an LLM model to obtain legal rule features;

[0135] The mathematical constraint generation module 906 is used to map the legal rule features to a preset mathematical space to generate mathematical constraints in the preset mathematical space.

[0136] The supplier data acquisition module 908 is used to acquire the initial suppliers at each level in the hierarchical supply chain, and to acquire the corresponding supplier data according to the level of each initial supplier.

[0137] The supplier data mapping module 910 is used to map supplier data to the preset mathematical space and verify that the supplier data meets the requirements of the mathematical constraints.

[0138] The first optimization module 912 is used to optimize the initial suppliers whose verification results do not meet the requirements of the mathematical constraints, so that the optimized hierarchical supply chain meets the requirements of the mathematical constraints.

[0139] In one embodiment, the legal text acquisition module 902 includes:

[0140] The structural decomposition submodule is used to decompose the specified product to obtain the BOM tree structure of the specified product; wherein, the BOM tree structure includes multiple nodes, and each node corresponds to a product component of the specified product;

[0141] The initial supplier acquisition submodule is used to acquire the initial suppliers of each node in the BOM tree structure and the importance of each node;

[0142] The hierarchical supply chain submodule is used to generate a hierarchical supply chain for the specified product based on the initial suppliers of each node and their corresponding importance.

[0143] In one embodiment, the legal text acquisition module 902 further includes:

[0144] The product component acquisition submodule is used to acquire the product components corresponding to each node.

[0145] The alternative supplier acquisition submodule is used to acquire multiple alternative suppliers corresponding to the node based on the product components.

[0146] A similarity score calculation submodule is used to calculate the similarity score between the initial supplier and each of the alternative suppliers for each node;

[0147] The sorting submodule is used to sort the alternative suppliers according to the size of the similarity scores.

[0148] In one embodiment, the legal text acquisition module 902 further includes:

[0149] The data information receiving submodule is used to receive the target function and the data information of each of the alternative suppliers;

[0150] The target supply chain acquisition submodule is used to input the various data information and the objective function into a preset mathematical space to optimize and solve the objective function to obtain the target supply chain;

[0151] The comparison information acquisition submodule is used to compare the target supply chain with the hierarchical supply chain to obtain comparison information.

[0152] The comparison information sending submodule is used to send the comparison information to a designated terminal.

[0153] In one embodiment, the legal rule feature acquisition module 904 includes:

[0154] The legal text segmentation submodule is used to perform hierarchical parsing of the legal text using a preset semantic segmentation technology, so as to segment the legal text into multiple paragraph modules;

[0155] The category attribute acquisition submodule is used to acquire the category attributes of each paragraph module and extract the paragraph module as clause paragraph module;

[0156] The three-dimensional rule feature matrix construction submodule is used to construct a three-dimensional rule feature matrix based on the clause paragraph module, which serves as the legal rule feature; wherein, the three-dimensional rule feature matrix includes an effectiveness dimension, a timeliness dimension, and a geographical dimension.

[0157] In one embodiment, the LLM-based cross-border supply chain optimization device further includes:

[0158] The specified information acquisition module is used to acquire specified information about the production location of the specified product and the location of the initial supplier.

[0159] The specified information mapping module is used to map the specified information to the preset mathematical space to generate the specified constraints of the preset mathematical space;

[0160] The verification module is used to map supplier data into the preset mathematical space and verify that the supplier data meets the requirements of the specified constraints.

[0161] The second optimization module is used to optimize the initial suppliers whose verification results do not meet the requirements of the specified constraints, so that the optimized initial suppliers meet the requirements of the specified constraints.

[0162] In one embodiment, the LLM-based cross-border supply chain optimization device further includes:

[0163] The production location acquisition module is used to acquire the locations of each target supplier in the optimized hierarchical supply chain, as well as the production location of the specified product;

[0164] The intelligent compliance certificate template acquisition module is used to acquire the corresponding intelligent compliance certificate template according to the production location and each of the locations;

[0165] The compliance data acquisition module is used to acquire the compliance data corresponding to each of the smart compliance certificate templates and fill it into the smart compliance certificate templates to obtain the smart compliance certificates corresponding to each target supplier.

[0166] Figure 4An internal structural diagram of an electronic device in one embodiment is shown. This electronic device can specifically be a terminal or a server, and more specifically, a computer device. Figure 4 As shown, the electronic device includes a processor, a memory, and a network interface connected via a system bus. The memory includes a non-volatile storage medium and internal memory. The non-volatile storage medium stores an operating system and may also store a computer program. When executed by the processor, this computer program enables the processor to implement an LLM-based cross-border supply chain optimization method. The internal memory may also store a computer program, which, when executed by the processor, enables the processor to implement an LLM-based cross-border supply chain optimization method. Those skilled in the art will understand that... Figure 4 The structure shown is merely a block diagram of a portion of the structure related to the present application and does not constitute a limitation on the electronic device to which the present application is applied. The specific electronic device may include more or fewer components than shown in the figure, or combine certain components, or have different component arrangements.

[0167] In one embodiment, an electronic device is provided, including a memory and a processor, the memory storing a computer program that, when executed by the processor, causes the processor to perform the following steps:

[0168] Obtain the hierarchical supply chain and related legal documents for the specified product;

[0169] The legal text is hierarchically parsed using an LLM model to obtain legal rule features;

[0170] The legal rule features are mapped to a preset mathematical space to generate mathematical constraints in the preset mathematical space;

[0171] Obtain the initial suppliers at each level of the hierarchical supply chain, and obtain the corresponding supplier data according to the level of each initial supplier;

[0172] The supplier data is mapped into the preset mathematical space, and the supplier data is verified to meet the requirements of the mathematical constraints.

[0173] The initial suppliers whose verification results do not meet the requirements of the mathematical constraints are optimized so that the optimized hierarchical supply chain meets the requirements of the mathematical constraints.

[0174] By performing hierarchical analysis of relevant legal texts, the system accurately extracts the characteristics of legal rules and maps them into a pre-defined mathematical space to form mathematical constraints. This automatically identifies and optimizes initial suppliers that do not meet these constraints, ensuring the compliance and stability of the entire supply chain. This reduces legal risks, enhances the transparency and flexibility of the supply chain, achieves cost savings and efficiency improvements, and enables the automated application of legal requirements to supply chain management, significantly improving the efficiency and accuracy of compliance audits.

[0175] In one embodiment, a computer-readable storage medium is provided storing a computer program that, when executed by a processor, causes the processor to perform the following steps:

[0176] Obtain the hierarchical supply chain and related legal documents for the specified product;

[0177] The legal text is hierarchically parsed using an LLM model to obtain legal rule features;

[0178] The legal rule features are mapped to a preset mathematical space to generate mathematical constraints in the preset mathematical space;

[0179] Obtain the initial suppliers at each level of the hierarchical supply chain, and obtain the corresponding supplier data according to the level of each initial supplier;

[0180] The supplier data is mapped into the preset mathematical space, and the supplier data is verified to meet the requirements of the mathematical constraints.

[0181] The initial suppliers whose verification results do not meet the requirements of the mathematical constraints are optimized so that the optimized hierarchical supply chain meets the requirements of the mathematical constraints.

[0182] By performing hierarchical analysis of relevant legal texts, the system accurately extracts the characteristics of legal rules and maps them into a pre-defined mathematical space to form mathematical constraints. This automatically identifies and optimizes initial suppliers that do not meet these constraints, ensuring the compliance and stability of the entire supply chain. This reduces legal risks, enhances the transparency and flexibility of the supply chain, achieves cost savings and efficiency improvements, and enables the automated application of legal requirements to supply chain management, significantly improving the efficiency and accuracy of compliance audits.

[0183] Those skilled in the art will understand that all or part of the processes in the above embodiments can be implemented by a computer program instructing related hardware. The program can be stored in a non-volatile computer-readable storage medium, and when executed, it can include the processes of the embodiments described above. Any references to memory, storage, databases, or other media used in the embodiments provided in this application can include non-volatile and / or volatile memory. Non-volatile memory can include read-only memory (ROM), programmable ROM (PROM), electrically programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM), or flash memory. Volatile memory can include random access memory (RAM) or external cache memory. By way of illustration and not limitation, RAM is available in various forms, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), dual data rate SDRAM (DDRSDRAM), enhanced SDRAM (ESDRAM), synchronous link DRAM (SLDRAM), RAMbus direct RAM (RDRAM), direct memory bus dynamic RAM (DRDRAM), and RAMbus dynamic RAM (RDRAM), etc.

[0184] The technical features of the above embodiments can be combined in any way. For the sake of brevity, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0185] The embodiments described above are merely illustrative of several implementation methods of this application, and while the descriptions are specific and detailed, they should not be construed as limiting the scope of this patent application. It should be noted that those skilled in the art can make various modifications and improvements without departing from the concept of this application, and these all fall within the protection scope of this application. Therefore, the protection scope of this patent application should be determined by the appended claims.

Claims

1. A cross-border supply chain optimization method based on LLM, characterized in that, The method includes: Obtain the hierarchical supply chain and related legal documents for the specified product; The legal text is hierarchically parsed using an LLM model to obtain legal rule features; The legal rule features are mapped to a preset mathematical space to generate mathematical constraints in the preset mathematical space; The initial suppliers at each level of the hierarchical supply chain are obtained, and the corresponding supplier data is obtained according to the level of each initial supplier; wherein, the supplier data includes the supplier's geographical location, main supplied products, price, delivery cycle, and compliance records; The supplier data is mapped into the preset mathematical space, and the supplier data is verified to meet the requirements of the mathematical constraints. The initial suppliers whose verification results do not meet the requirements of the mathematical constraints are optimized so that the optimized hierarchical supply chain meets the requirements of the mathematical constraints. The step of obtaining the hierarchical supply chain for the specified product includes: The specified product is structurally decomposed to obtain the BOM tree structure of the specified product; wherein the BOM tree structure includes multiple nodes, and each node corresponds to a product component of the specified product; Obtain the initial suppliers and importance of each node in the BOM tree structure; A hierarchical supply chain for the specified product is generated based on the initial suppliers of each node and their corresponding importance. Obtain the product components corresponding to each node; Based on the product composition, obtain multiple alternative suppliers corresponding to the node; Receive the objective function and data information from each of the alternative suppliers; The data and objective function are input into a preset mathematical space to optimize and solve the objective function, thereby obtaining the target supply chain. By comparing the target supply chain with the hierarchical supply chain, comparison information is obtained; The comparison information is sent to the designated terminal.

2. The LLM-based cross-border supply chain optimization method according to claim 1, characterized in that, The step of performing hierarchical parsing of the legal text using an LLM model to obtain legal rule features includes: The legal text is hierarchically parsed using a pre-defined semantic segmentation technique to divide it into multiple paragraph modules; Obtain the category attributes of each paragraph module and extract the paragraph modules as clause paragraph modules; A three-dimensional rule feature matrix is ​​constructed based on the aforementioned clause paragraph module to serve as the legal rule feature; wherein, the three-dimensional rule feature matrix includes a validity dimension, a timeliness dimension, and a geographical dimension.

3. The LLM-based cross-border supply chain optimization method according to claim 1, characterized in that, After the steps of obtaining the initial suppliers at each level of the hierarchical supply chain and obtaining the corresponding supplier data according to the level of each initial supplier, the method further includes: Obtain the specified information regarding the production location of the designated product and the location of the initial supplier; The specified information is mapped to the preset mathematical space to generate the specified constraints of the preset mathematical space; The supplier data is mapped into the preset mathematical space, and the supplier data is verified to meet the specified constraints. The initial suppliers whose verification results show that they do not meet the requirements of the specified constraints are optimized so that the optimized initial suppliers meet the requirements of the specified constraints.

4. The LLM-based cross-border supply chain optimization method according to claim 1, characterized in that, After the step of optimizing the initial suppliers whose verification results do not meet the requirements of the mathematical constraints, so that the optimized hierarchical supply chain meets the requirements of the mathematical constraints, the method further includes: Obtain the locations of each target supplier in the optimized hierarchical supply chain, as well as the production location of the specified product; Obtain the corresponding smart compliance certificate template based on the production location and each of the aforementioned locations; Obtain the compliance data corresponding to each of the smart compliance certificate templates and fill it into the smart compliance certificate templates to obtain the smart compliance certificates corresponding to each target supplier.

5. A cross-border supply chain optimization device based on LLM, characterized in that, The device includes: The legal document acquisition module is used to acquire the hierarchical supply chain and related legal documents for a specified product. The legal rule feature acquisition module is used to perform hierarchical parsing of the legal text using an LLM model to obtain legal rule features; A mathematical constraint generation module is used to map the legal rule features to a preset mathematical space to generate mathematical constraints in the preset mathematical space. The supplier data acquisition module is used to acquire the initial suppliers at each level of the hierarchical supply chain, and to acquire the corresponding supplier data according to the level of each initial supplier; wherein, the supplier data includes the supplier's geographical location, main supplied products, price, delivery cycle, and compliance records; The supplier data mapping module is used to map supplier data into the preset mathematical space and verify that the supplier data meets the requirements of the mathematical constraints. The first optimization module is used to optimize the initial suppliers whose verification results do not meet the requirements of the mathematical constraints, so that the optimized hierarchical supply chain meets the requirements of the mathematical constraints. The legal text acquisition module includes: The structural decomposition submodule is used to decompose the specified product to obtain the BOM tree structure of the specified product; wherein, the BOM tree structure includes multiple nodes, and each node corresponds to a product component of the specified product; The initial supplier acquisition submodule is used to acquire the initial suppliers of each node in the BOM tree structure and the importance of each node; The hierarchical supply chain submodule is used to generate a hierarchical supply chain for the specified product based on the initial suppliers of each node and their corresponding importance. The product component acquisition submodule is used to acquire the product components corresponding to each node. The alternative supplier acquisition submodule is used to acquire multiple alternative suppliers corresponding to the node based on the product components. The data information receiving submodule is used to receive the target function and the data information of each of the alternative suppliers; The target supply chain acquisition submodule is used to input the various data information and the objective function into a preset mathematical space to optimize and solve the objective function to obtain the target supply chain; The comparison information acquisition submodule is used to compare the target supply chain with the hierarchical supply chain to obtain comparison information. The comparison information sending submodule is used to send the comparison information to a designated terminal.

6. A computer-readable storage medium, characterized in that, The system contains a computer program that, when executed by a processor, causes the processor to perform the steps of the LLM-based cross-border supply chain optimization method as described in any one of claims 1 to 4.

7. An electronic device, characterized in that, The device includes a memory and a processor, the memory storing a computer program that, when executed by the processor, causes the processor to perform the steps of the LLM-based cross-border supply chain optimization method as described in any one of claims 1 to 4.

Images