Coal purchasing decision method, system and device based on large language model

Abstract

The present application relates to the technical field of coal purchasing decision support, and specifically discloses a coal purchasing decision method, system and device based on a large language model, the coal purchasing decision method based on the large language model comprising: obtaining target data; inputting the target data into the large language model to output a coal purchasing basic data analysis result; according to the coal purchasing basic data analysis result, further analyzing through the large language model to generate a fuel purchasing strategy direction suggestion; according to a strategy direction selected by a user, constructing a multi-objective optimization model and solving the same to obtain an optimal combination of a purchased coal type, quantity and purchasing route; and according to the optimal combination of the purchased coal type, quantity and route, generating a customized coal purchasing scheme through the natural language generation capability of the large language model. The embodiments of the present application realize the intelligentization, scientification and high efficiency of coal purchasing decision, and provide strong support for reducing costs, improving efficiency and ensuring safety for thermal power enterprises.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of coal purchasing decision support, and in particular to a coal purchasing decision method, system and equipment based on a large language model. BACKGROUND

[0002] The main challenge of current coal purchasing decision is excessive reliance on artificial experience judgment, which is difficult to achieve multi-objective collaborative optimization. The traditional decision-making mode has three limitations:

[0003] Firstly, it is highly subjective and cannot quantitatively evaluate the mutual influence of key factors such as coal structure, fuel cost, transportation cycle, environmental protection indicators, and inventory structure. Currently, the coal purchasing of thermal power enterprises is basically based on artificial experience decision-making, which is based on future power generation, environmental protection, inventory, and available coal types, and combined with Excel for manual calculation. The coal types to be purchased are considered to be qualified purchases under the requirements of average calorific value and sulfur. This method is difficult to ensure the optimal combination of the number of coal types to be purchased when facing many coal types to be purchased. Even in the case of a certain number of coal types, manual calculation cannot guarantee the optimal combination of the purchase quantity of each single coal type.

[0004] Secondly, it has a lagging response and is difficult to respond to dynamic changes in coal prices, transportation capacity, and weather. Coal purchasing and coal blending are complementary, and scientific coal purchasing is the source of ensuring the effectiveness of future coal blending. If the distribution of the purchased coal types is unreasonable, it will inevitably cause problems such as not being able to carry high load, environmental protection exceeding standards, and high-calorific-value coal carrying low load. However, artificial decision-making cannot track market changes in real time, resulting in delayed adjustment of purchasing strategies.

[0005] Thirdly, the strategy dimension is single, and it is difficult to balance economy and environmental protection while ensuring supply. Artificial purchasing can only provide a single purchasing scheme, and cannot provide a combined purchasing scheme. The selectability and rapidity of the purchasing scheme cannot be guaranteed. Traditional methods usually only aim to minimize the purchasing cost, while ignoring the comprehensive optimization of multi-dimensional factors such as combustion efficiency, environmental compliance, and transportation efficiency.

[0006] Currently, the related technologies for coal purchasing optimization mainly include the following categories:

[0007] Purchasing guidance model based on operational optimization model: such as the fuel purchasing guidance model based on large CFB boiler disclosed in Chinese patent CN108846538A, which achieves a certain degree of optimization, but lacks flexibility in handling complex constraint conditions, and does not integrate the latest artificial intelligence technology.

[0008] Virtual blending-based procurement optimization system: A virtual blending-based cascade coal procurement optimization system is disclosed in Chinese Patent CN202410797510. Although the coal blending factor is considered, the natural language processing and knowledge reasoning capabilities of large language models are not fully utilized.

[0009] Large model-based procurement application: ChatGPT is used in coal enterprise procurement management, mainly for procurement demand prediction, price analysis and evaluation, compliance risk management, etc. Although large language models are introduced, they are not deeply integrated with professional operational optimization models, making it difficult to achieve multi-objective optimization of complex procurement decisions.

[0010] Existing technology: Lack of multi-objective collaborative optimization capability, usually only targeting a single objective such as minimizing procurement cost, making it difficult to simultaneously consider economic efficiency, safety, efficiency, environmental protection, transportation cycle, inventory ratio, and other multiple objectives. Limited intelligence, lack of natural language interaction capabilities and knowledge reasoning capabilities, unable to efficiently collaborate with users in human-machine decision-making. Insufficient data integration and processing capabilities, difficult to effectively integrate multi-source heterogeneous data, including equipment parameters, inventory information, supplier information, coal prices, transportation routes, and other complex data. Poor dynamic adaptability, difficult to respond to dynamic factors such as coal price fluctuations, transportation capacity changes, and weather conditions in real time, lacking the ability to quickly respond to market changes. SUMMARY

[0011] Therefore, it is necessary to provide a large language model-based coal procurement decision method, system and device to address the above technical problems. By integrating large language models and operational optimization models, the coal procurement decision-making process is made intelligent, scientific and efficient, providing strong support for power generation enterprises to reduce costs, improve efficiency, and ensure safety.

[0012] In a first aspect, a large language model-based coal procurement decision method is provided, comprising the following steps:

[0013] Obtain target data, including boiler equipment design parameters, coal minimum inventory indicators, fuel procurement plan development basis information, coal supplier and price information, historical procurement coal information, coal transportation route and cycle information;

[0014] Input the target data into a pre-trained large language model, and analyze it through the large language model according to the preset prompt words, outputting the coal procurement basic data analysis result;

[0015] According to the coal procurement basic data analysis result, policies and regulations, and industry standards, further analyze through the large language model to generate fuel procurement strategy direction suggestions;

[0016] According to the strategy direction selected by the user from the fuel procurement strategy direction suggestion, combined with the planned procurement quantity output by the basic operation and the high and low ash fusion point expected procurement proportion, a multi-objective optimization model is constructed, and the multi-objective optimization model is solved to obtain the optimal combination of procurement coal type, procurement quantity and procurement route.

[0017] According to the optimal combination of procurement coal type, procurement quantity and procurement route, a customized coal procurement plan is generated through the natural language generation capability of the large language model.

[0018] In some examples, the analysis by the large language model outputs a coal procurement basic data analysis result, including:

[0019] The target data is verified and cleaned to generate a data verification report.

[0020] Key information is extracted from the verified and cleaned target data and converted into structured data.

[0021] The association between different data items in the structured data is analyzed to obtain a preliminary procurement decision.

[0022] According to the abnormal data items in the data verification report, the influence on the procurement decision is evaluated, and the coal procurement basic data analysis result is obtained according to the evaluation result.

[0023] In some examples, the large language model further analyzes the coal procurement basic data analysis result, policies and regulations and industry standards to generate a fuel procurement strategy direction suggestion, including:

[0024] Obtain environmental protection policies, energy policies and coal industry standard information to analyze the policy constraints of the procurement environment.

[0025] According to the type of power plant and the unit capacity parameter, obtain the industry standard to obtain a procurement strategy that meets the industry specifications.

[0026] Analyze historical procurement data and current market information to predict market trends in coal price trends and supply tightness.

[0027] According to the policy constraints of the procurement environment, the procurement strategy that meets the industry specifications, and the market trends of coal price trends and supply tightness, generate multiple fuel procurement strategy direction suggestions.

[0028] In some examples, according to the strategy direction selected by the user from the fuel procurement strategy direction suggestion, combined with the planned procurement quantity output by the basic operation and the high and low ash fusion point expected procurement proportion, a multi-objective optimization model is constructed, and the multi-objective optimization model is solved to obtain the optimal combination of procurement coal type, procurement quantity and procurement route, including:

[0029] According to the power generation plan, the power supply coal consumption index information, the total planned procurement quantity, and the predicted procurement proportion basis data of high and low ash melting point coal types are calculated;

[0030] According to multiple fuel procurement strategy direction suggestions, define the decision variables of the large language model;

[0031] According to the user-selected fuel procurement strategy direction suggestion, build the objective function;

[0032] According to the boiler equipment parameters, inventory requirements, and environmental protection standard factors, set the constraint conditions of the large language model;

[0033] Solve the large language model to generate the optimal combination of procurement coal types, procurement quantities, and procurement routes.

[0034] In some examples, according to the optimal combination of procurement coal types, procurement quantities, and procurement routes, a customized coal procurement plan is generated through the natural language generation capability of the large language model, including:

[0035] According to the solution results of the multi-objective optimization model, a detailed data table is generated, wherein the detailed data table includes the procurement quantity, supplier, price, transportation mode, and arrival time of each coal type;

[0036] According to the preset prompt words, a structured procurement report is generated, including the current coal procurement and inventory situation, next month's demand situation, and next month's fuel procurement plan;

[0037] Key data in the coal procurement plan is displayed in the form of charts and tables;

[0038] The generated detailed data table and structured procurement report are output to a specified path and uploaded to the enterprise's document management system.

[0039] In some examples, it further includes:

[0040] Through a natural language interaction interface, real-time interaction with the user is realized to receive the user's evaluation, adjustment, and confirmation of the generated coal procurement plan, and the parameters and strategy suggestions of the large language model are optimized according to the user feedback.

[0041] In some examples, the natural language interaction interface is used to realize real-time interaction with the user to receive the user's evaluation, adjustment, and confirmation of the generated coal procurement plan, and the parameters and strategy suggestions of the large language model are optimized according to the user feedback, including:

[0042] A text or voice-based interaction interface is provided to have a conversation with the user through natural language;

[0043] Analyzing the natural language text input by the user to identify the user's true needs and intentions;

[0044] Adjusting and optimizing the coal procurement plan according to the user's feedback and needs.

[0045] In some examples, further comprising:

[0046] Querying professional knowledge related to coal procurement through the interactive interface and providing accurate answers based on external knowledge bases;

[0047] Recording the user's interaction history and decision preferences, and continuously optimizing the parameters and strategy recommendations of the large language model through continuous learning.

[0048] In a second aspect, a large language model-based coal procurement decision system is provided, comprising:

[0049] An acquisition module for acquiring target data, including boiler equipment design parameters, coal type minimum inventory indicators, fuel procurement plan development basis information, coal supplier and price information, historical procurement coal information, coal transportation route and cycle information;

[0050] An analysis module for inputting the target data into a pre-trained large language model and analyzing through the large language model according to a preset prompt word to output coal procurement basic data analysis results;

[0051] A generation module for further analyzing through the large language model based on the coal procurement basic data analysis results, policies and regulations, and industry standards to generate fuel procurement strategy direction recommendations;

[0052] A solving module for constructing a multi-objective optimization model based on the total planned procurement quantity and high-low ash melting point estimated procurement proportion output by the basic operation, and solving the multi-objective optimization model to obtain the optimal combination of procurement coal type, procurement quantity, and procurement route according to the strategy direction selected by the user from the fuel procurement strategy direction recommendations;

[0053] A decision module for generating a customized coal procurement plan through the natural language generation capability of the large language model based on the optimal combination of procurement coal type, procurement quantity, and procurement route.

[0054] In a third aspect, a computer device is provided, comprising a processor and a computer program, wherein the processor executes the computer program to implement the large language model-based coal procurement decision method according to the first aspect described above.

[0055] By integrating large language models and operations research optimization models, embodiments of the present invention have achieved intelligent, scientific, and efficient coal procurement decisions, providing strong support for thermal power companies to reduce costs, improve efficiency, and ensure safety. Attached Figure Description

[0056] Other features, objects, and advantages of the invention will become more apparent from the following detailed description of non-limiting embodiments with reference to the accompanying drawings:

[0057] Figure 1 A flowchart illustrating a coal procurement decision-making method based on a large language model, as provided in one embodiment of the present invention;

[0058] Figure 2 A flowchart of a coal procurement decision-making method based on a large language model, provided as another embodiment of the present invention;

[0059] Figure 3 This is a detailed flowchart illustrating the construction and solution of the multi-objective optimization model of the present invention;

[0060] Figure 4 A structural block diagram of a coal procurement decision-making system based on a large language model provided in an embodiment of the present invention;

[0061] Figure 5 A structural block diagram of a computer device provided in an embodiment of the present invention. Detailed Implementation

[0062] The present invention will now be described in further detail with reference to the embodiments and accompanying drawings. It should be understood that the specific embodiments described herein are for illustrative purposes only and are not intended to limit the invention. Furthermore, it should be noted that, for ease of description, only the parts relevant to the invention are shown in the accompanying drawings.

[0063] It should be noted that, unless otherwise specified, the embodiments and features of the embodiments in this invention can be combined with each other. The invention will now be described in detail with reference to the accompanying drawings and embodiments.

[0064] The following describes in detail, with reference to the accompanying drawings, a coal procurement decision-making method, system, and equipment based on a large language model according to embodiments of the present invention.

[0065] Figure 1 This is a flowchart of a coal procurement decision-making method based on a large language model according to an embodiment of the present invention. Figure 1 As shown, and in combination Figure 2 A coal procurement decision-making method based on a large language model according to an embodiment of the present invention includes the following steps:

[0066] S101: Obtain target data, including boiler equipment design parameters, coal minimum inventory indicators, fuel procurement plan development basis information, coal supplier and price information, historical procurement coal information, coal transportation routes and cycle information.

[0067] S102: Input the target data into the pre-trained large language model, and analyze the target data through the large language model according to a preset prompt word to output a coal-fired procurement basis data analysis result.

[0068] In an embodiment of the present application, the analysis through the large language model to output the coal-fired procurement basis data analysis result includes: checking and cleaning the target data to generate a data checking report; extracting key information from the checked and cleaned target data and converting the key information into structured data; analyzing the correlation between different data items in the structured data to obtain a preliminary procurement decision; evaluating the influence on the procurement decision according to the abnormal data items of the data checking report, and obtaining the coal-fired procurement basis data analysis result according to the evaluation result.

[0069] S103: According to the coal-fired procurement basis data analysis result, policy and regulation and industry standard, further analyze through the large language model to generate a fuel procurement strategy direction suggestion.

[0070] In an embodiment of the present application, according to the coal-fired procurement basis data analysis result, policy and regulation and industry standard, further analyze through the large language model to generate a fuel procurement strategy direction suggestion, including: obtaining environmental protection policy, energy policy and coal industry standard information to analyze the policy constraints of the procurement environment; obtaining industry standards according to power plant types and unit capacity parameters to obtain procurement strategies that meet industry standards; analyzing historical procurement data and current market information to predict coal price trends and market trends of supply tightness; generating multiple fuel procurement strategy direction suggestions according to the policy constraints of the procurement environment, the procurement strategies that meet the industry standards, and the market trends of the coal price trends and the supply tightness.

[0071] S104: According to the strategy direction selected by the user from the fuel procurement strategy direction suggestions, combine the planned procurement quantity output by the basis operation and the high-low ash melting point expected procurement proportion to construct a multi-objective optimization model, and solve the multi-objective optimization model to obtain an optimal combination of procurement coal types, procurement quantities and procurement routes.

[0072] In an embodiment of the present application, according to the strategy direction selected by the user from the fuel procurement strategy direction suggestion, a multi-objective optimization model is constructed in combination with the planned procurement quantity output by the basic operation, the predicted procurement proportion of high and low ash melting point, and the multi-objective optimization model is solved to obtain the optimal combination of procurement coal type, procurement quantity and procurement route, including: calculating the planned procurement quantity, the predicted procurement proportion of high and low ash melting point coal type based on the power generation plan, the power supply coal consumption index information; defining the decision variables of the large language model according to the multiple fuel procurement strategy direction suggestions; constructing the objective function according to the user-selected fuel procurement strategy direction suggestion; setting the constraint conditions of the large language model according to the boiler equipment parameters, inventory requirements, environmental protection standard factors; solving the large language model to generate the optimal combination of procurement coal type, procurement quantity and procurement route.

[0073] S105: According to the optimal combination of procurement coal type, procurement quantity and procurement route, a customized coal procurement plan is generated through the natural language generation capability of the large language model.

[0074] In an embodiment of the present application, according to the optimal combination of procurement coal type, procurement quantity and procurement route, a customized coal procurement plan is generated through the natural language generation capability of the large language model, including: generating a detailed data table according to the solving result of the multi-objective optimization model, wherein the detailed data table includes the procurement quantity, supplier, price, transportation mode and arrival time of each coal type; generating a structured procurement report according to the preset prompt words, wherein the structured procurement report includes the current coal procurement and inventory situation, the next month's demand situation and the next month's fuel procurement plan; displaying the key data in the coal procurement plan in the form of charts; outputting the generated detailed data table and structured procurement report to the specified path and uploading them to the enterprise's document management system.

[0075] In an embodiment of the present application, the coal procurement decision-making method based on the large language model further includes: realizing real-time interaction with the user through a natural language interaction interface to receive the user's evaluation, adjustment and confirmation of the generated coal procurement plan, and optimizing the parameters and strategy suggestions of the large language model according to the user feedback.

[0076] In this example, the natural language interaction interface is provided to realize real-time interaction with the user to receive the user's evaluation, adjustment and confirmation of the generated coal procurement plan, and optimize the parameters and strategy suggestions of the large language model according to the user feedback, including: providing a text or voice-based interaction interface to have a conversation with the user through natural language; analyzing the natural language text input by the user to identify the user's real needs and intentions; adjusting and optimizing the coal procurement plan according to the user's feedback and needs;

[0077] The coal purchasing decision-making method based on a large language model, characterized in that it further comprises: querying professional knowledge related to coal purchasing through an interactive interface, and providing accurate answers based on an external knowledge base; recording the interaction history and decision preferences of the user, and continuously optimizing the parameters and strategy recommendations of the large language model through continuous learning.

[0078] For example, in combination with Figure 3 As shown in the embodiment, data acquisition and integration acquire various types of data related to coal purchasing through multiple channels, including:

[0079] 1. Equipment design parameters: Obtain key parameters such as boiler model, rated evaporation capacity, designed coal type, and burner type from the equipment management system of the power plant, which will be used to evaluate the adaptability of different coal types to unit operation.

[0080] 2. Minimum inventory index: Obtain the minimum inventory requirements of various coal types from the fuel management system to ensure that the procurement plan can meet the safety inventory demand.

[0081] 3. Information for developing fuel procurement plans: Obtain monthly power generation plans, power supply coal consumption indicators, and blending schemes from the production planning department to provide a basis for developing procurement plans.

[0082] 4. Supplier information: Obtain basic information, supply capacity, quality assurance system, and historical performance of each coal supplier from the supplier management system for supplier evaluation and selection.

[0083] 5. Coal price information: Through data interfaces with coal trading platforms and price index publishing agencies, real-time market price information of various coal types is obtained, including pit price, plant price, and tax-included price in different pricing methods.

[0084] 6. Transportation line and cycle information: Obtain transportation line, transportation mode (railway, highway, waterway), transportation cycle, and transportation cost information from the logistics management system for optimizing transportation schemes.

[0085] 7. Historical procurement coal type information: Obtain procurement records of the past 12 months from the procurement management system, including procurement coal type, quantity, price, supplier, and transportation mode, for analyzing procurement trends and predicting market changes.

[0086] These data interact with external systems through a data interface module to ensure real-time and accurate data. The data integration module converts these multi-source heterogeneous data into a unified structured format and stores them in the database of the system for subsequent analysis.

[0087] In this embodiment, the integrated data is analyzed using a pre-trained large language model, such as the DeepSeek-R1 model, as follows:

[0088] Data verification and cleaning: The large model verifies the input data according to the preset prompt "You are an expert in purchasing fuel coal for power plants. Please summarize and analyze the verification results of the above data from the perspective of fuel coal procurement, and analyze the impact of the data items with abnormal verification results." It identifies missing values, outliers, and inconsistent data, and generates a data verification report. For example, the large model may find that the calorific value data of coal provided by a supplier deviates by more than 10% from historical records, and mark it as an abnormal data item.

[0089] Data feature extraction: The large model extracts key information from unstructured data (such as supplier qualification documents and transportation contract texts) through natural language processing techniques, such as supplier credit rating and transportation contract terms, and converts them into structured data for subsequent analysis.

[0090] Data correlation analysis: The large model analyzes the correlation between different data items, such as the correlation between coal prices and quality indicators such as calorific value and sulfur content, to identify coal types with abnormal prices; analyzes the relationship between transportation costs and transportation distance and mode, and finds the most economical transportation solution.

[0091] Abnormal data impact assessment: For abnormal data items found in the verification, the large model assesses their potential impact on procurement decisions. For example, if the sulfur content data of a batch of coal is abnormally high, the large model will analyze the potential environmental compliance risks and increased desulfurization costs, and provide a prompt in the data analysis results.

[0092] The output of the large model data analysis is a coal procurement basic data analysis report, including data quality assessment, key data features, abnormal data and their impact, etc., providing a basis for subsequent procurement strategy suggestion generation.

[0093] In this embodiment, the procurement strategy suggestion generation module generates multiple procurement strategy direction suggestions based on the data analysis results of the large model, combined with information such as policies and industry standards in the external knowledge base, as follows:

[0094] Policy and regulation compliance analysis: The large model obtains the latest environmental protection policies, energy policies, and coal industry standards from the external knowledge base, and analyzes the policy constraints of the current procurement environment. For example, if the latest environmental protection policy requires that the sulfur content of fuel coal for power plants not exceed 0.8%, the large model will incorporate this requirement into the constraints of the procurement strategy.

[0095] Industry Standard Matching: Based on the type of power plant and unit capacity, the large model obtains the corresponding industry standards from external knowledge bases, such as the "Guidelines for Coal-fired Power Plants" and other relevant standards, to ensure that the procurement strategy complies with industry norms.

[0096] Market Trend Analysis: The large model analyzes historical procurement data and current market information to predict coal price trends, supply tightness, and other market trends, providing a reference for procurement strategies. For example, if the large model predicts that the price of a certain coal type will rise by 10% in the next month, it will recommend increasing the procurement quantity of that coal type in advance.

[0097] Strategy Direction Generation: Based on the above analysis, the large model generates multiple procurement strategy direction recommendations, including:

[0098] Optimal Unit Energy Consumption (Maximize Heat Efficiency): Prioritize coal types with high calorific value and good combustion performance, even if the price is relatively high, to improve unit heat efficiency and reduce coal consumption.

[0099] Optimal Environmental Indicators (Ensure Emission Compliance): Prioritize clean coal types with low sulfur and ash content to meet environmental requirements and reduce pollutant emissions, which may increase procurement costs but reduce environmental treatment costs.

[0100] Shortest Supply Cycle (Optimal Transportation Route and Shortest Transportation Cycle): Prioritize suppliers with short transportation distances and reliable transportation methods to ensure timely coal supply and reduce inventory pressure, which may increase transportation costs.

[0101] Optimal Inventory Structure (Optimal Proportion of High and Low Ash Melting Point Inventory): Determine the optimal proportion of high and low ash melting point coal types based on unit operation requirements and inventory status to optimize inventory structure and improve inventory turnover efficiency.

[0102] The large model generates strategy recommendations based on the preset prompt "Fuel Procurement Strategy Direction Recommendations: You are a fuel coal procurement expert for power plants. Based on the above data, recommend a procurement direction in the following categories: Lowest Price of In-plant Coal Standard Coal Unit Price (Focus on Cost Optimization), Optimal Unit Energy Consumption (Maximize Heat Efficiency), Optimal Environmental Indicators (Ensure Emission Compliance), Shortest Supply Cycle (Improve Supply Chain Response), and Optimal Inventory Structure (Optimize Inventory Structure), and provide an explanation (only the reason needs to be given, no recommendations need to be provided)." The advantages and disadvantages of each strategy direction are analyzed to provide users with options.

[0103] In this embodiment, the multi-objective optimization model construction module constructs a multi-objective optimization model according to the user-selected strategy direction, in combination with the planned total procurement quantity, the estimated procurement proportion of high and low ash fusion point, and other information output by the basic operation. The specific steps are as follows:

[0104] Basic operation: According to the power generation plan, the power supply coal consumption index, and other information, calculate the planned total procurement quantity, the estimated procurement proportion of high and low ash fusion point coal, and other basic data. For example, according to the monthly power generation plan and the power supply coal consumption index, the total standard coal required can be calculated, and then according to the historical blending data and the unit characteristics, the approximate proportion of high and low ash fusion point coal can be determined.

[0105] Decision variable definition: According to the procurement strategy direction, define the decision variables of the model. For example, in the cost optimization strategy, the decision variables may include the procurement quantity of each coal type, the supplier selection, the transportation mode selection, etc.; in the environmental protection index optimization strategy, the decision variables may include the sulfur content, the ash content, and other quality indicators of each coal type.

[0106] Objective function construction: According to the user-selected strategy direction, construct the corresponding objective function. For example: Cost optimization objective function: minimize the in-plant coal standard coal unit price, while considering procurement cost, transportation cost, and inventory cost, etc.

[0107] Thermal efficiency maximization objective function: maximize the unit thermal efficiency, mainly considering the calorific value, volatile content, and other combustion characteristic indicators of coal types.

[0108] Environmental protection index optimization objective function: minimize pollutant emissions, mainly considering the sulfur content, ash content, nitrogen content, and other indicators of coal types.

[0109] Constraint condition setting: According to equipment parameters, inventory requirements, environmental protection standards, and other factors, set the constraint conditions of the model, including:

[0110] Supply capacity constraint: the maximum supply capacity limit of each supplier.

[0111] Quality index constraint: the calorific value, sulfur content, ash content, and other quality indicators of coal types must meet the requirements of unit operation and environmental protection.

[0112] Transportation capacity constraint: the maximum transportation capacity limit of the transportation line.

[0113] Inventory constraint: the procurement quantity must meet the minimum inventory requirement, while not exceeding the maximum inventory capacity.

[0114] Policy and regulation constraint: must comply with the latest environmental protection policies, energy policies, and other requirements.

[0115] Multi-objective optimization solution: use appropriate multi-objective optimization algorithm to solve the model, generate a set of optimal solutions, represent the trade-off relationship between different objectives. Users can choose the most satisfactory solution from the optimal solution according to actual needs.

[0116] The coal purchasing decision method based on large language model of the embodiment of the present application can balance multiple objectives such as economy, safety, energy efficiency, environmental protection, transportation cycle, and inventory ratio at the same time, generate a purchasing plan with optimal comprehensive indicators, and avoid the limitations of single objective optimization in traditional methods. Through the combination of large language model and operational optimization model, an end-to-end solution from natural language requirements to mathematical model solving is realized. Through the natural language understanding and knowledge reasoning ability of the large language model, intelligent decision support is provided, which can understand complex user requirements and generate purchasing strategy suggestions that meet the requirements. The natural language interaction interface greatly improves the usability and interaction efficiency of the system. It can effectively integrate multi-source heterogeneous data, including device parameters, inventory information, supplier information, coal prices, transportation routes, and other complex data, and extract key information through the analysis ability of the large model to provide comprehensive data support for decision-making. It can real-time perceive market changes, transportation fluctuations, weather conditions and other dynamic factors, quickly adjust the purchasing strategy, and ensure that the purchasing plan is always in the optimal state. Through the interface with external data systems, real-time data updating and dynamic model adjustment are realized.

[0117] In addition, a rich external knowledge base is integrated, including policies and regulations, industry standards, coal type information, unit information, etc., providing comprehensive knowledge support for decision-making and improving the scientificity and compliance of decision-making. The knowledge reasoning ability of the large model can apply these knowledge to specific purchasing scenarios and provide more targeted suggestions. Through automated data processing and model solving, the workload of manual analysis and calculation is greatly reduced, and the decision-making efficiency is improved. At the same time, the precise solution of the multi-objective optimization model and the intelligent analysis of the large model ensure the quality and accuracy of the decision-making. By optimizing the purchasing combination, transportation scheme and inventory structure, the purchasing cost and life cycle cost are effectively reduced. Through comprehensive analysis of factors such as supplier risk, market risk, environmental risk, potential risks are identified and evaluated in advance, risk response strategies are provided, and the risk control ability of enterprises is enhanced.

[0118] In summary, the coal purchasing decision method based on large language model of the embodiment of the present application according to the embodiment of the present application integrates large language model and operational optimization model, realizes the intelligentization, scientization and high efficiency of coal purchasing decision, and provides strong support for thermal power enterprises to reduce cost, improve efficiency and ensure safety.

[0119] Figure 4 is the structure diagram of the coal purchasing decision system based on large language model according to an embodiment of the present application. AsFigure 4 As shown, the coal purchasing decision-making system based on a large language model according to one embodiment of the application comprises an acquisition module 410, an analysis module 420, a generation module 430, a solving module 440, and a decision-making module 450, wherein:

[0120] The acquisition module 410 is configured to acquire target data, which includes boiler equipment design parameters, minimum inventory indicators of coal types, fuel purchasing plan formulation basis information, coal supplier and price information, historical purchasing coal type information, and coal transportation routes and cycle information.

[0121] The analysis module 420 is configured to input the target data into a pre-trained large language model, and analyze the large language model according to a preset prompt word to output a coal purchasing basis data analysis result.

[0122] The generation module 430 is configured to further analyze the large language model according to the coal purchasing basis data analysis result, policies and regulations, and industry standards to generate a fuel purchasing strategy direction suggestion.

[0123] The solving module 440 is configured to combine the plan purchasing total amount and the high and low ash melting point predicted purchasing proportion output by the basis operation according to a strategy direction selected by a user from the fuel purchasing strategy direction suggestion, construct a multi-objective optimization model, and solve the multi-objective optimization model to obtain an optimal combination of a purchasing coal type, a purchasing quantity, and a purchasing route.

[0124] The decision-making module 450 is configured to generate a customized coal purchasing scheme through the natural language generation capability of the large language model according to the optimal combination of the purchasing coal type, the purchasing quantity, and the purchasing route.

[0125] Specifically, the data acquisition is responsible for acquiring data related to coal purchasing from multiple data sources, including:

[0126] An equipment parameter interface is connected with an equipment management system of a power plant to acquire equipment design parameters such as a boiler model, a rated evaporation capacity, a design coal type, and a burner type.

[0127] An inventory management interface is connected with a fuel management system to acquire minimum inventory indicators and current inventory data of various coal types.

[0128] A production plan interface is connected with a system of a production plan department to acquire monthly power generation capacity plans, power supply coal consumption indicators, blending schemes, and other fuel purchasing plan formulation basis information.

[0129] A supplier management interface is connected with a supplier management system to acquire basic information, supply capacity, quality assurance systems, and historical performance of various coal suppliers.

[0130] Price Information Interface: Connects with coal trading platforms and price index publishing agencies to obtain real-time market price information for various coal types.

[0131] Logistics Management Interface: Connects with logistics management systems to obtain information such as transportation routes, transportation methods, transportation cycles, and transportation costs from coal production areas to power plants.

[0132] Purchase History Interface: Connects with purchase management systems to obtain purchase records from the past 12 months, including coal types, quantities, prices, suppliers, and transportation methods.

[0133] These interfaces can be Web service interfaces based on standard protocols such as HTTP / REST, SOAP, or direct data access interfaces based on database connections, ensuring efficient data acquisition and real-time updates.

[0134] Data Integration converts heterogeneous data from multiple data sources into a unified structured format for subsequent processing, including:

[0135] Data Cleaning Component: Cleans the acquired data to remove noise, handle missing values and outliers, and improve data quality.

[0136] Data Conversion Component: Converts data in different formats and encodings into a unified structured format such as JSON or XML.

[0137] Data Integration Component: Integrates data from different data sources into a unified data model, addressing data conflicts and inconsistencies.

[0138] Data Storage Component: Stores the integrated data in the system's database for subsequent module use. The database can use a relational database (such as MySQL, Oracle) or a NoSQL database (such as MongoDB), depending on data characteristics and processing needs.

[0139] Large Model Analysis is responsible for analyzing and processing integrated data, including:

[0140] Large Language Model: Uses a pre-trained large language model such as DeepSeek-R1 model, with strong natural language understanding, knowledge reasoning and generation capabilities.

[0141] Prompt Management Component: Manages and maintains prompts for different tasks, such as data verification prompts, strategy suggestion prompts, and report generation prompts, ensuring that the large model can accurately understand user requirements and generate outputs that meet requirements.

[0142] Model Invocation Interface: Provides a unified interface for other modules to call the functions of the large model, such as data analysis, strategy suggestion generation, report generation, etc.

[0143] Model Output Processing Component: Processes and analyzes the output results of the large model, extracts key information and converts it into structured data for subsequent modules.

[0144] The large model analysis module can be deployed on high-performance computing servers, using GPU acceleration to improve processing efficiency and ensure that the system can respond to user requests in real time.

[0145] The external knowledge base stores various types of knowledge related to coal procurement, providing background knowledge support for large model analysis, including:

[0146] Policy and Regulation Knowledge Base: Stores the latest environmental protection policies, energy policies, coal industry standards, etc., to ensure the compliance of procurement decisions.

[0147] Industry Standard Knowledge Base: Stores industry standards such as "Guidelines for Coal Quality for Coal-fired Power Plants" to provide a basis for coal selection and quality assessment.

[0148] Coal Type Characteristics Knowledge Base: Stores the physical and chemical properties, combustion characteristics, and applicable scenarios of various coal types to help assess the adaptability of different coal types to unit operation.

[0149] Supplier Evaluation Knowledge Base: Stores supplier evaluation indicators, evaluation methods, risk factors, etc., to provide references for supplier selection.

[0150] Transportation Knowledge Graph: Constructs a knowledge graph of coal transportation networks, including transportation routes, transportation methods, transportation costs, transportation times, etc., to support transportation scheme optimization.

[0151] The external knowledge base module can store knowledge in various forms such as knowledge graph, relational database, or document database to ensure efficient retrieval and application of knowledge.

[0152] Multi-objective optimization is responsible for constructing and solving multi-objective optimization models to generate the optimal procurement scheme, including:

[0153] Model Construction Component: According to the user's selected strategy direction and basic data, construct the corresponding multi-objective optimization model, define the decision variables, objective function and constraint conditions.

[0154] Algorithm Library: Integrates multiple multi-objective optimization algorithms, selects the appropriate algorithm for solving according to the problem characteristics.

[0155] Model Solving Component: Uses the selected algorithm to solve the model, generates a set of optimal solutions for the user to choose from.

[0156] Result analysis component: analyze and evaluate the solution results, provide performance indicators, risk analysis, sensitivity analysis and other information of the solution, help users make decisions.

[0157] Multi-objective optimization module can be realized by MATLAB, Python and other mathematical modeling tools, or a special optimization solving engine can be developed to ensure the accuracy and efficiency of the model.

[0158] The scheme generation generates customized procurement schemes according to the solution results of the multi-objective optimization model, including:

[0159] Excel generation component: convert the solution results into detailed data tables in Excel format, including the procurement quantity, supplier, price, transportation mode, arrival time and other detailed information of each coal type.

[0160] Word generation component: use the natural language generation capability of large models to generate structured procurement reports, including current coal procurement and inventory, next month's demand, next month's fuel procurement plan and other content.

[0161] Chart generation: display key data in the procurement scheme in chart form, such as bar chart, line chart, scatter chart, etc., to enhance the visualization effect of the scheme.

[0162] Scheme output: output the generated Excel table and Word report to the specified path and automatically upload to the enterprise's document management system for easy access and use by relevant departments.

[0163] The scheme generation module can integrate Microsoft Office or wps components or use open source office document generation libraries (such as Python's openpyxl, python-docx library) to ensure that the generated documents meet the enterprise standards.

[0164] Interactive interface: provides an interface for users to interact with the system, including:

[0165] Natural language interaction component: provides a text or voice-based interactive interface that allows users to communicate with the system through natural language.

[0166] User intent understanding component: analyzes the user's natural language text input, identifies the user's real needs and intentions, and converts them into instructions that the system can understand.

[0167] Scheme display component: displays the generated procurement scheme in an intuitive way, including tables, charts, reports and other forms, to help users understand and compare the pros and cons of different schemes.

[0168] User Feedback Collection Component: Collects user evaluations, suggestions, and adjustment requirements for the generated plan, providing basis for system optimization.

[0169] Knowledge Query Component: Supports user queries related to professional knowledge such as coal type characteristics, environmental protection policies, market dynamics, etc., and provides accurate answers.

[0170] The interactive interface module can be implemented in various forms such as Web applications, desktop applications, or mobile applications, ensuring that users can conveniently use the system on different devices.

[0171] Data docking is responsible for data exchange and synchronization with external data systems, including:

[0172] Data Lake Docking Component: Interfaces with the enterprise's data lake (ERP, CRM, etc.) to obtain and synchronize related data.

[0173] Weather Docking Component: Interfaces with weather systems to obtain weather data and analyze the impact of weather conditions on coal production, transportation, and inventory.

[0174] Plant-side Fuel Intelligence System Docking Component: Interfaces with the plant-side fuel intelligence system to obtain real-time information such as coal quality data and inventory data, supporting dynamic optimization of procurement plans.

[0175] Data Synchronization Strategy Component: Defines data synchronization strategies and frequencies to ensure that data in the system is consistent with external systems, improving data real-time and accuracy.

[0176] The data docking module can be implemented using ETL tools, data integration platforms, or custom-developed data interfaces to ensure efficient docking with external systems.

[0177] The following is an example of a specific application scenario to illustrate the practical application process of the invention.

[0178] A thermal power plant has four 300MW coal-fired generating units, with a design coal type of Jinbei bituminous coal. The current inventory is 150,000 tons, and the minimum inventory requirement is 100,000 tons. According to the monthly production plan, the planned power generation for next month is 360 million kilowatt-hours, and the target power supply coal consumption is 300 grams per kilowatt-hour. Currently, the coal market price fluctuates greatly, and the environmental protection department requires the implementation of new emission standards starting next month, with a sulfur content of no more than 0.8% for coal-fired power plants. The power plant needs to develop a coal procurement plan for next month to ensure that production requirements and environmental protection requirements are met while controlling procurement costs.

[0179] 1. Data Acquisition and Integration

[0180] Equipment parameter interface obtains information such as boiler type and design coal type.

[0181] The inventory management interface obtains the current inventory of coal and the minimum inventory requirement.

[0182] The production plan interface obtains the monthly power generation plan and the power supply coal consumption target.

[0183] The price information interface obtains the market prices of various coal types, including Jinbei bituminous coal (Qnet,ar=5500kcal / kg, S=0.6%, delivered price 650 yuan / ton), Shaanxi coal (Qnet,ar=5000kcal / kg, S=0.7%, delivered price 580 yuan / ton), Inner Mongolia coal (Qnet,ar=5800kcal / kg, S=0.9%, delivered price 680 yuan / ton), etc.

[0184] The logistics management interface obtains the transportation cycle and transportation cost of each coal type.

[0185] The procurement history interface obtains the procurement records of the past three months to analyze procurement trends.

[0186] 2. Large model data analysis

[0187] The large model verifies the obtained data and finds that the sulfur content of Inner Mongolia coal (0.9%) exceeds the new environmental protection standard (0.8%), marking it as an abnormal data item and evaluating its impact on procurement decisions.

[0188] The cost-effectiveness of each coal type is analyzed, finding that Shaanxi coal has a lower price but also lower calorific value, Inner Mongolia coal has high calorific value but exceeds the sulfur standard, and Jinbei bituminous coal meets environmental protection requirements but has a higher price.

[0189] The future one-month coal price trend is predicted, finding that due to production restrictions in the production area, the price of Jinbei bituminous coal may rise by 5%.

[0190] 3. Generation of procurement strategy suggestions:

[0191] Based on the results of data analysis, the large model generates three procurement strategy direction suggestions:

[0192] Cost optimization strategy: preferentially select Shaanxi coal with lower prices, but when mixing high-sulfur coal, increase desulfurization costs.

[0193] Environmental protection index optimization strategy: select Jinbei bituminous coal with sulfur content meeting requirements, but may face the risk of rising prices.

[0194] Comprehensive balance strategy: mix purchasing Jinbei bituminous coal and Shaanxi coal to meet environmental protection requirements and control procurement costs.

[0195] The large model recommends the comprehensive balance strategy, reasoning that it can achieve a good balance between environmental compliance and cost control, while considering the risk of future price fluctuations.

[0196] 4. Multi-objective optimization model construction:

[0197] The standard coal required for the next month is 108,000 tons (3.6 billion kWh x 300g / kWh ÷ 1000).

[0198] Define decision variables as the purchase quantity of Jinbei bituminous coal (x1), Shaanxi coal (x2), and Inner Mongolia coal (x3).

[0199] Construct a multi-objective optimization model:

[0200] Minimize procurement cost: 650x1 + 580x2 + 680x3

[0201] Minimize sulfur content: (0.6x1 + 0.7x2 + 0.9x3) / (x1 + x2 + x3) ≤ 0.8%

[0202] Maximize average calorific value: (5500x1 + 5000x2 + 5800x3) / (x1 + x2 + x3) ≥ 5300 kcal / kg

[0203] Set constraints:

[0204] x1 + x2 + x3 ≥ 108,000 tons (meet production needs)

[0205] x1 + x2 + x3 ≤ 150,000 tons (not exceed storage capacity)

[0206] x3 = 0 (Inner Mongolia coal sulfur exceeds standard, prohibited procurement)

[0207] x1 ≥ 20,000 tons (ensure a certain proportion of design coal, ensure combustion stability)

[0208] 5. Procurement scheme generation and output:

[0209] The optimal solution obtained by solving the multi-objective optimization model is x1 = 40,000 tons, x2 = 68,000 tons, procurement cost is 6,544 million yuan, average sulfur content is 0.72%, and average calorific value is 5200 kcal / kg.

[0210] Generate an Excel table detailing procurement quantities, suppliers, prices, transportation methods, and other information.

[0211] Generate a Word report containing current inventory, next month's demand analysis, recommended procurement scheme, and its advantages and risks.

[0212] 6. System interaction and feedback:

[0213] User views the generated procurement scheme through the interactive interface and asks, "Can we increase the proportion of Shaanxi coal to reduce costs?"

[0214] The system re-runs the multi-objective optimization model, adjusts the constraints, and generates a new solution: x1=30,000 tons, x2=78,000 tons, with a procurement cost of 64.44 million yuan, an average sulfur content of 0.75%, and an average calorific value of 5100 kcal / kg.

[0215] After comparing the two options, the user chooses the second option. The system records the user's preference and optimizes the model parameters.

[0216] Through this invention, power plants successfully developed procurement plans that balanced environmental requirements with cost control, reducing procurement costs by approximately 1.5% while ensuring that the sulfur content of coal met new environmental standards. The system's intelligent analysis and multi-objective optimization functions help users quickly assess the impact of different strategies, improving decision-making efficiency and quality.

[0217] The coal procurement decision-making system based on a large language model, as described in this invention, can simultaneously balance multiple objectives such as economy, safety, energy efficiency, environmental protection, transportation cycle, and inventory ratio, generating a procurement plan with optimal comprehensive indicators, thus avoiding the limitations of single-objective optimization in traditional methods. By combining a large language model with an operations research optimization model, an end-to-end solution is achieved, from natural language requirements to mathematical model solutions. The natural language understanding and knowledge reasoning capabilities of the large language model provide intelligent decision support, understanding complex user needs and generating suitable procurement strategy suggestions. The natural language interface greatly improves the system's usability and interaction efficiency. It can effectively integrate multi-source heterogeneous data, including complex data such as equipment parameters, inventory information, supplier information, coal prices, and transportation routes, and extract key information through the analytical capabilities of the large model, providing comprehensive data support for decision-making. It can perceive dynamic factors such as market changes, transportation capacity fluctuations, and weather conditions in real time, quickly adjusting procurement strategies to ensure the procurement plan is always in an optimal state. Through integration with external data systems, real-time data updates and dynamic model adjustments are achieved.

[0218] In summary, the coal procurement decision-making system based on a large language model according to embodiments of the present invention, by integrating a large language model with an operations research optimization model, achieves intelligent, scientific, and efficient coal procurement decision-making, providing strong support for thermal power enterprises to reduce costs, improve efficiency, and ensure safety.

[0219] The following is for reference. Figure 5 , Figure 5 A schematic diagram of a computer device structure suitable for implementing embodiments of the present invention is shown.

[0220] like Figure 5As shown, the computer system 1000 includes a central processing unit (CPU) 1001 which can perform various appropriate actions and processes in accordance with a program stored in a read only memory (ROM) 1002 or a program loaded from the storage section 1008 into a random access memory (RAM) 1003. In the RAM 1003, various programs and data necessary for the operation of the system are also stored. The CPU 1001, the ROM 1002, and the RAM 1003 are connected to each other through a bus 1004. An input / output (I / O) interface 1005 is also connected to the bus 1004.

[0221] Connected to the I / O interface 1005 are an input section 1006 including a keyboard, a mouse, etc.; an output section 1007 including a display such as a cathode ray tube (CRT), a liquid crystal display (LCD), etc., and a speaker, etc.; a storage section 1008 including a hard disk, etc.; and a communication section 1009 including a network interface card such as a LAN card, a modem, etc. The communication section 1009 performs communication processing via a network such as the Internet. A drive 1010 is also connected to the I / O interface 1005 as necessary. A removable media 1011 such as a magnetic disk, an optical disk, a magneto-optical disk, a semiconductor memory, etc. is attached to the drive 1010 as necessary, so that a computer program read therefrom is installed into the storage section 1008 as necessary.

[0222] In particular, the processes described above, with reference to the flowcharts Figure 1 may be implemented as a computer program. For example, an embodiment of the present application includes a computer-readable storage medium comprising a computer program which contains program code for performing the processes illustrated in the flowcharts Figure 1 of the methods shown.

[0223] In particular, the processes described above, with reference to the flowcharts Figure 1 may be implemented as a computer program. For example, an embodiment of the present application includes a computer-readable storage medium comprising a computer program which contains program code for performing the processes illustrated in the flowcharts Figure 1 of the methods shown.

[0224] In such an embodiment, the computer program contains program code for performing the methods illustrated in the flowcharts. In such an embodiment, the computer program can be downloaded and installed from a network via the communication section 1009, and / or installed from the removable media 1011. When the computer program is executed by the central processing unit (CPU) 1001, the above-described functions defined in the system of the present application are performed.

[0225] The above embodiments only express several implementation manners of the present application, and the description is relatively specific and detailed, but it should not be understood as a limitation on the patent scope of the application. It should be noted that for ordinary skilled persons in the art, several modifications and improvements can be made without departing from the concept of the present application, which are all within the protection scope of the present application. Therefore, the protection scope of the patent of the present application should be subject to the appended claims.

Claims

1. A coal purchasing decision-making method based on a large language model, characterized in that, The method comprises the following steps: obtaining target data, the target data comprising boiler equipment design parameters, coal type minimum inventory index, fuel procurement plan formulation basis information, coal supplier and price information, historical procurement coal type information, coal transportation route and cycle information; inputting the target data into a pre-trained large language model, and analyzing the target data through the large language model according to a preset prompt word to output a coal procurement basis data analysis result; further analyzing the coal procurement basis data analysis result, policies and regulations and industry standards through the large language model to generate a fuel procurement strategy direction suggestion; according to a strategy direction selected by a user from the fuel procurement strategy direction suggestion, combining a planned procurement quantity output by a basis operation and a high-low ash melting point predicted procurement proportion, constructing a multi-objective optimization model, and solving the multi-objective optimization model to obtain an optimal combination of procurement coal type, procurement quantity and procurement route, comprising: calculating a planned procurement quantity and a high-low ash melting point coal type predicted procurement proportion basis data according to a power generation capacity plan and a power supply coal consumption index information; defining a decision variable of the large language model according to a plurality of fuel procurement strategy direction suggestions; constructing an objective function according to a fuel procurement strategy direction suggestion selected by the user, the objective function comprising a cost optimization objective function, a thermal efficiency maximization objective function and an environmental protection index optimization objective function, wherein the cost optimization objective function is used for minimizing an incoming plant coal standard coal unit price, the thermal efficiency maximization objective function is used for maximizing a unit thermal efficiency, and the environmental protection index optimization objective function is used for minimizing a pollutant emission; setting a constraint condition of the multi-objective optimization model according to boiler equipment parameters, inventory requirements and environmental protection standard factors; and solving the multi-objective optimization model by using an operational optimization algorithm to generate an optimal combination of procurement coal type, procurement quantity and procurement route; generating a customized coal procurement scheme through a natural language generation capability of the large language model according to the optimal combination of procurement coal type, procurement quantity and procurement route.

2. The coal procurement decision-making method based on a large language model according to claim 1, characterized in that, The analysis through the large language model and the output of the coal procurement basis data analysis result comprise: verifying and cleaning the target data to generate a data verification report; extracting key information from the verified and cleaned target data and converting the key information into structured data; analyzing the correlation between different data items in the structured data to obtain a preliminary procurement decision; evaluating the influence on the procurement decision according to an abnormal data item of the data verification report, and obtaining a coal procurement basis data analysis result according to an evaluation result.

3. The coal procurement decision-making method based on a large language model according to claim 1, characterized in that, The further analysis through the large language model and the generation of the fuel procurement strategy direction suggestion according to the coal procurement basis data analysis result, policies and regulations and industry standards comprise: obtaining environmental protection policies, energy policies and coal industry standard information to analyze policy constraints of a procurement environment; obtaining industry standards according to power plant types and unit capacity parameters to obtain a procurement strategy conforming to industry specifications; analyzing historical procurement data and current market information to predict coal price trends and market trends of supply tightness; According to the policy constraints of the procurement environment, the procurement strategy conforming to the industry standards, and the market trends of coal price trends and supply tightness, multiple fuel procurement strategy direction suggestions are generated.

4. The coal procurement decision-making method based on a large language model according to claim 1, characterized in that, According to the optimal combination of the procurement coal type, the procurement quantity and the procurement route, a customized coal procurement plan is generated through the natural language generation capability of the large language model, including: According to the solution results of the multi-objective optimization model, a detailed data table is generated, wherein the detailed data table includes the procurement quantity, supplier, price, transportation mode and arrival time of each coal type. According to the preset prompt words, a structured procurement report is generated, including the current coal procurement and inventory situation, the next month's demand situation, and the next month's fuel procurement plan. The key data in the coal procurement plan is displayed in the form of charts. The generated detailed data table and structured procurement report are output to the specified path and uploaded to the enterprise's document management system.

5. The coal procurement decision-making method based on a large language model according to claim 1, characterized in that, Further comprising: Through the natural language interaction interface, real-time interaction with the user is realized to receive the user's evaluation, adjustment and confirmation of the generated coal procurement plan, and the parameters and strategy suggestions of the large language model are optimized according to the user feedback.

6. The coal procurement decision-making method based on a large language model according to claim 5, characterized in that, The natural language interaction interface is used to realize real-time interaction with the user to receive the user's evaluation, adjustment and confirmation of the generated coal procurement plan, and the parameters and strategy suggestions of the large language model are optimized according to the user feedback, including: Providing a text or voice-based interaction interface to have a conversation with the user through natural language; Analyzing the natural language text input by the user to identify the user's real needs and intentions; Adjust and optimize the coal procurement plan according to the user's feedback and needs.

7. The coal procurement decision-making method based on a large language model according to claim 6, characterized in that, Further comprising: Querying professional knowledge related to coal procurement through the interaction interface and providing accurate answers based on external knowledge base; Record the user's interaction history and decision preferences, and continuously optimize the parameters and strategy suggestions of the large language model through continuous learning.

8. A coal purchasing decision system based on a large language model, characterized by, Including: An acquisition module for acquiring target data, including boiler equipment design parameters, coal type minimum inventory indicators, fuel procurement plan development basis information, coal supplier and price information, historical procurement coal type information, coal transportation route and cycle information; An analysis module for inputting the target data into a pre-trained large language model and analyzing the data through the large language model according to preset prompt words to output coal procurement basic data analysis results; A generation module for generating fuel procurement strategy direction suggestions through further analysis of the large language model based on the coal procurement basic data analysis results, policies and regulations, and industry standards; The solving module is configured to, according to a strategy direction selected by a user from the fuel procurement strategy direction suggestions, combine a planned procurement quantity output by a basic operation and a predicted procurement proportion of high and low ash melting point to construct a multi-objective optimization model, and solve the multi-objective optimization model to obtain an optimal combination of a procurement coal type, a procurement quantity and a procurement route, including: calculating planned procurement quantity and predicted procurement proportion of high and low ash melting point coal based on basic data according to power generation plan and power supply coal consumption index information; defining decision variables of the large language model according to a plurality of fuel procurement strategy direction suggestions; constructing an objective function according to a fuel procurement strategy direction suggestion selected by a user, the objective function including a cost optimization objective function, a thermal efficiency maximization objective function and an environmental protection index optimization objective function, wherein the cost optimization objective function is used to minimize the in-plant coal standard coal unit price, the thermal efficiency maximization objective function is used to maximize the unit thermal efficiency, and the environmental protection index optimization objective function is used to minimize pollutant emissions; setting constraint conditions of the multi-objective optimization model according to boiler equipment parameters, inventory requirements and environmental protection standard factors; and solving the multi-objective optimization model by using an operational optimization algorithm to generate an optimal combination of a procurement coal type, a procurement quantity and a procurement route. The decision module is configured to generate a customized coal procurement scheme by using the natural language generation capability of the large language model according to the optimal combination of the procurement coal type, the procurement quantity and the procurement route. 9.A computer device, comprising a memory and a processor, wherein a computer program is stored on the memory, and the computer device is characterized in that, The processor executes the computer program to implement the large language model-based coal procurement decision method according to any one of claims 1-7.

Images