Fuel optimization scheduling method based on coal supply chain prediction

Abstract

The invention provides a fuel optimization scheduling method based on coal supply chain prediction, and belongs to the technical field of supply chain management, and the method comprises the steps: collecting the related historical data of a coal supply chain based on a to-be-realized target of fuel optimization, and carrying out the supply analysis and demand analysis of the related historical data of the coal supply chain, obtaining a supply data set and a demand data set so as to obtain supply features and demand features, performing feature fusion on the supply features and the demand features to form a comprehensive feature set, and constructing a prediction model about the supply-demand relationship of the coal supply chain based on the comprehensive feature set; according to the prediction result, constructing a scheduling model, outputting a fuel scheduling scheme, and performing fuel scheduling to obtain an actual scheduling result; the actual scheduling result is compared with the prediction result to obtain a supply-demand comparison result, and the prediction model is optimized based on the supply-demand comparison result, so that the problems of information lag, inaccurate prediction, inflexible scheduling and the like are effectively solved, and the adaptability and efficiency of the supply chain are enhanced.

Description

Technical Field

[0001] The present invention relates to the technical field of supply chain management, and in particular to a fuel optimization scheduling method based on coal supply chain prediction. Background Art

[0002] In the coal supply chain, the fluctuation of energy demand and the complexity of the supply chain make accurate supply and demand forecasting and scheduling the key. Traditional supply chain management technologies often have problems such as information lag, insufficient forecast accuracy and inflexible scheduling. Similarly, rigid scheduling strategies are difficult to cope with sudden demand or supply changes, limiting the flexibility and responsiveness of the supply chain.

[0003] Therefore, the present invention provides a fuel optimization scheduling method based on coal supply chain prediction. Summary of the invention

[0004] The present invention provides a fuel optimization scheduling method based on coal supply chain forecasting. The method collects and analyzes historical data of the coal supply chain, extracts supply and demand characteristics, and integrates and constructs a forecasting model to improve the accuracy and flexibility of supply and demand forecasting. The optimization model is fed back by actual results to achieve more dynamic scheduling decisions, improve the response speed of the supply chain, effectively solve the problems of information lag, inaccurate forecasting and inflexible scheduling, and enhance the adaptability and efficiency of the supply chain.

[0005] The present invention provides a fuel optimization scheduling method based on coal supply chain forecasting, comprising: Step 1: Based on the goals to be achieved in fuel optimization, relevant historical data on the coal supply chain are collected, and supply analysis and demand analysis are performed on the relevant historical data of the coal supply chain to obtain a supply data set and a demand data set; Step 2: extracting supply features from the supply data set, extracting demand features from the demand data set, fusing the supply features with the demand features to form a comprehensive feature set, and constructing a prediction model for the supply and demand relationship of the coal supply chain based on the comprehensive feature set; Step 3: Based on the prediction results output by the prediction model, a scheduling model is constructed, a fuel scheduling plan is output, fuel scheduling is performed, and actual scheduling results are obtained; Step 4: Compare the actual scheduling results with the forecast results to obtain the supply and demand comparison results, and optimize the forecast model based on the supply and demand comparison results.

[0006] The present invention provides a fuel optimization scheduling method based on coal supply chain prediction, which collects relevant historical data about the coal supply chain based on the target to be achieved in fuel optimization, and performs supply analysis and demand analysis on the relevant historical data of the coal supply chain to obtain a supply data set and a demand data set, including: Obtaining the target to be achieved for fuel optimization, collecting relevant historical data based on the target to be achieved for fuel optimization, and classifying the relevant historical data of the coal supply chain to obtain external historical data and internal historical data; Performing supply analysis on external historical data to obtain a first supply set, performing supply analysis on internal historical data to obtain a second supply set, and combining the first supply set and the second supply set to obtain a supply data set; A demand analysis is performed on external historical data to obtain a first demand set, a demand analysis is performed on an internal historical data set to obtain a second demand set, and the first demand set and the second demand set are combined to obtain a demand data set.

[0007] The present invention provides a fuel optimization scheduling method based on coal supply chain prediction, extracting supply features from the supply data set, extracting demand features from the demand data set, fusing the supply features with the demand features to form a comprehensive feature set, and constructing a prediction model for the supply and demand relationship of the coal supply chain based on the comprehensive feature set, including: Performing data structure analysis on the supply data set to determine a first key field of the supply data set, extracting the first key field as a first supply feature, performing feature derivation based on the first supply feature to obtain a second supply feature, and combining the first supply feature and the second supply feature to determine a supply feature set; Performing data structure analysis on the demand data set to determine a second key field of the demand data set, extracting the second key field as a first demand feature, performing feature derivation based on the first demand feature to obtain a second demand feature, and combining the first demand feature and the second demand feature to determine a demand feature set; A logical analysis is performed on the goals to be achieved in fuel optimization to derive the target logic, the demand characteristics and supply characteristics are combined based on the target logic to form a comprehensive feature set, and a prediction model for the supply and demand relationship of the coal supply chain is constructed based on the comprehensive feature set.

[0008] The present invention provides a fuel optimization scheduling method based on coal supply chain prediction, which constructs a scheduling model to output a fuel scheduling plan according to the prediction results output by the prediction model, and performs fuel scheduling, including: Extract the forecast results of fuel demand from the forecast model, identify the relevant factors affecting the scheduling process based on the goals to be achieved in fuel optimization, determine the constraints in the scheduling process based on the relevant factors, and select the optimization algorithm to build the scheduling model according to the forecast results and the constraints; A fuel scheduling plan is output according to the scheduling model, and the fuel scheduling plan is executed to obtain an actual scheduling result.

[0009] The present invention provides a fuel optimization scheduling method based on coal supply chain prediction, selecting an optimization algorithm according to the prediction results and constraint conditions, and constructing a scheduling model, including: Determine the objective function based on the target to be achieved in fuel optimization, randomly generate initial solutions according to the constraint conditions and the prediction results, calculate the objective function value of each initial solution, and set a dynamic threshold to make an excellent judgment on the objective function value of each initial solution to obtain an excellent solution; Extracting the excellent features of the excellent solution to set the disturbance range, and performing local search for the excellent solution based on the disturbance range to generate multiple candidate solutions, calculating the objective function values ​​of the candidate solutions, and judging the objective function values ​​of the candidate solutions according to the dynamic threshold to obtain the final solution; A scheduling model is constructed based on the final solution.

[0010] The present invention provides a fuel optimization scheduling method based on coal supply chain prediction, and sets a dynamic threshold to perform excellent judgment on the objective function value, including: ,in, Indicates at time t Dynamic threshold of Indicates a point in history The objective function value function of Indicates the current time point t The objective function value function of Indicates the current time point t Distribution function of the objective function value; Indicates the current time point t The dynamic threshold at the previous time point; Indicates a point in history; t Indicates the current time point; Indicates time period Neidi i1 The objective function value of the initial solution; Indicates time period Neidi i2 The objective function value of the initial solution; k1 Indicates from The total number of time steps back through time; k2 Indicates from t The total number of time steps back through time; j1 Indicates from Going back in time j1 step; j2 Indicates from t Going back in time j2 step; Indicates time point The number of initial solutions at time t; Indicates time point The number of initial solutions at time t; Represents the adjustment factor of the mean change of the objective function value; Indicates the influence factor of the objective function value at the historical time point on the objective function value at the current time point; Represents the rate factor of change of objective function value; Represents the distribution function change adjustment factor; Indicates the acceleration adjustment factor of the distribution function change; When the objective function value of the initial solution at the same time point is higher than the dynamic threshold, the corresponding initial solution is an excellent solution.

[0011] The present invention provides a fuel optimization scheduling method based on coal supply chain prediction, compares the actual scheduling result with the prediction result, obtains the supply and demand comparison result, and optimizes the prediction model based on the supply and demand comparison result, including: Perform error analysis on the actual scheduling results and the predicted results to determine the error situation at each stage of the fuel scheduling plan, and obtain a supply and demand comparison result based on the error situation; Conduct regularity analysis on the supply and demand comparison results to determine potential regularities of supply and demand changes, and optimize the prediction model based on the potential regularities of supply and demand changes.

[0012] The present invention provides a fuel optimization scheduling method based on coal supply chain forecasting, which performs regularity analysis on the supply and demand comparison results, determines the potential regularity of supply and demand changes, and optimizes the forecasting model based on the potential regularity of supply and demand changes, including: Identify abnormalities in the supply and demand comparison results, determine abnormal points, and trace the causes of the abnormal points. Based on the abnormal points and causes, summarize the potential patterns of supply and demand changes; The potential rules and anomalies are incorporated into the prediction model to optimize the prediction model.

[0013] Compared with the existing technology, the beneficial effects of this application are as follows: by collecting and analyzing historical data of the coal supply chain, extracting supply and demand characteristics, and integrating them to build a prediction model to improve the accuracy and flexibility of supply and demand forecasts, the optimization model is optimized through actual result feedback to achieve more dynamic scheduling decisions, improve the response speed of the supply chain, effectively solve problems such as information lag, inaccurate prediction and inflexible scheduling, and enhance the adaptability and efficiency of the supply chain.

[0014] Other features and advantages of the present invention will be described in the following description, and partly become apparent from the description, or understood by practicing the present invention. The purpose and other advantages of the present invention can be realized and obtained by the structures particularly pointed out in the written description and the accompanying drawings.

[0015] The technical solution of the present invention is further described in detail below through the accompanying drawings and embodiments. BRIEF DESCRIPTION OF THE DRAWINGS

[0016] The accompanying drawings are used to provide a further understanding of the present invention and constitute a part of the specification. Together with the embodiments of the present invention, they are used to explain the present invention and do not constitute a limitation of the present invention. In the accompanying drawings: Figure 1 It is a flow chart of a fuel optimization scheduling method based on coal supply chain prediction provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0017] The preferred embodiments of the present invention are described below in conjunction with the accompanying drawings. It should be understood that the preferred embodiments described herein are only used to illustrate and explain the present invention, and are not used to limit the present invention.

[0018] Embodiment 1: The embodiment of the present invention provides a fuel optimization scheduling method based on coal supply chain prediction, such as Figure 1 As shown, including: Step 1: Based on the goals to be achieved in fuel optimization, relevant historical data on the coal supply chain are collected, and supply analysis and demand analysis are performed on the relevant historical data of the coal supply chain to obtain a supply data set and a demand data set; Step 2: extracting supply features from the supply data set, extracting demand features from the demand data set, fusing the supply features with the demand features to form a comprehensive feature set, and constructing a prediction model for the supply and demand relationship of the coal supply chain based on the comprehensive feature set; Step 3: Based on the prediction results output by the prediction model, a scheduling model is constructed, a fuel scheduling plan is output, fuel scheduling is performed, and actual scheduling results are obtained; Step 4: Compare the actual scheduling results with the forecast results to obtain the supply and demand comparison results, and optimize the forecast model based on the supply and demand comparison results.

[0019] In this embodiment, the goal to be achieved refers to the specific goal that is hoped to be achieved in the process of coal supply chain optimization, such as cost minimization: reducing the total cost of coal procurement, transportation and storage; efficiency maximization: improving the efficiency of coal transportation and distribution, and reducing idle and waiting time.

[0020] In this embodiment, relevant historical data refers to historical information that needs to be analyzed during the optimization process, such as transportation cost data: cost records of different transportation methods in the past year; inventory level data: monthly coal inventory and its changes.

[0021] In this embodiment, supply analysis is an evaluation of supply data with the goal of understanding the capacity and efficiency of the supply chain, such as external supplier evaluation: analyzing the delivery time and quality of different external suppliers; transportation efficiency analysis: evaluating the cost and time of different transportation routes.

[0022] In this embodiment, demand analysis is an evaluation of demand data with the goal of understanding the patterns and changes in customer demand, such as demand trend analysis: analyzing seasonal demand changes over the past few years; customer preference analysis: identifying demand characteristics and changes of major customers.

[0023] In this embodiment, the supply data set is a comprehensive data set formed by combining the first supply set and the second supply set, for example, comprehensive supply information: integrated data of quotations from external suppliers and production capacity of internal coal mines.

[0024] In this embodiment, the demand data set is a comprehensive data set formed by combining the first demand set and the second demand set, for example, comprehensive demand information: integrated data of external market demand and internal customer demand.

[0025] In this embodiment, the process of data fusion is to analyze the supply data set to determine the first key field (such as supply quantity, supply time) and extract it as the first supply feature; analyze the demand data set to determine the second key field (such as demand quantity, demand time) and extract it as the first demand feature; in the feature derivation stage, extract relevant information (such as supply volatility) based on the first supply feature to form the second supply feature; extract relevant information (such as demand growth rate) based on the first demand feature to form the second demand feature; and construct a complete set of supply and demand features.

[0026] In this embodiment, the input of the prediction model is all the features in the comprehensive feature set, such as "customer-supplier matching", "demand volatility", "supply stability", etc.; the output of the prediction model refers to the result generated by the model, such as the forecast of coal demand in the future; the training set is a feature set containing historical demand and supply data, which is used to train the model to identify patterns and relationships.

[0027] In this embodiment, the supply and demand relationship of the coal supply chain refers to the interaction and dependency between suppliers (such as coal mines, producers) and demanders (such as power plants, industrial users) during the process of coal production, transportation, storage and consumption. For example, the supplier: production capacity: the mining capacity and production efficiency of coal mines, transportation capacity: the transportation method and efficiency from coal mines to consumption sites, inventory management: coal storage and inventory levels to ensure the continuity of supply.

[0028] In this embodiment, the forecast results include supply forecast results, demand forecast results, and supply balance forecast results. For example, the supply balance forecast result is that in the next week, the total demand is expected to be 600 tons, while the total supply is 450 tons, and there is a shortage of 150 tons. If the inventory is 200 tons, it is expected that there may be a risk of insufficient supply during peak demand periods (such as Wednesdays).

[0029] In this embodiment, the input of the scheduling model is the fuel demand forecast results, relevant factors, constraints, current inventory levels, transportation capacity, etc., for example, the current inventory (such as 300 tons), the demand forecast for the next week (such as 100 tons per day); the output is the specific arrangement of the fuel scheduling plan, such as the daily delivery volume, transportation route, delivery time, etc. For example, 100 tons of fuel are scheduled to be delivered every day, divided into two deliveries, one in the morning and the other in the afternoon.

[0030] In this embodiment, the fuel scheduling plan is a specific scheduling plan, including the time, quantity and method of delivery, for example, 100 tons are delivered every day from Monday to Friday, and the delivery volume is adjusted according to changes in demand on weekends.

[0031] In this embodiment, the actual scheduling result is the actual amount of fuel delivered, time and any deviation after the execution of the scheduling plan. For example, the actual delivery is 95 tons delivered on Monday and 105 tons delivered on Tuesday, and deviations are recorded (such as delayed delivery time due to traffic delays).

[0032] In this embodiment, the supply and demand comparison result is the difference between the actual supply and demand obtained based on the error analysis. For example, if in a certain period of time, the actual supply is 1,100 tons and the actual demand is 1,200 tons, the supply and demand comparison result is that the supply-demand gap is 100 tons.

[0033] In this embodiment, anomalies are identified on the supply and demand comparison results, abnormal points are determined and their causes are traced, and the potential laws of supply and demand changes are summarized, and these laws and abnormal points are incorporated into the prediction model to optimize the model parameters.

[0034] The working principle and beneficial effects of the above technical solution are: by collecting and analyzing historical data of the coal supply chain, extracting supply and demand characteristics, and integrating them to build a prediction model to improve the accuracy and flexibility of supply and demand forecasts, and optimizing the model through actual result feedback, more dynamic scheduling decisions can be made, and the response speed of the supply chain can be improved. It effectively solves problems such as information lag, inaccurate predictions and inflexible scheduling, and enhances the adaptability and efficiency of the supply chain.

[0035] Embodiment 2: The embodiment of the present invention provides a fuel optimization scheduling method based on coal supply chain prediction, which collects relevant historical data about the coal supply chain based on the target to be achieved in fuel optimization, and performs supply analysis and demand analysis on the relevant historical data of the coal supply chain to obtain a supply data set and a demand data set, including: Obtaining the target to be achieved for fuel optimization, collecting relevant historical data based on the target to be achieved for fuel optimization, and classifying the relevant historical data of the coal supply chain to obtain external historical data and internal historical data; Performing supply analysis on external historical data to obtain a first supply set, performing supply analysis on internal historical data to obtain a second supply set, and combining the first supply set and the second supply set to obtain a supply data set; A demand analysis is performed on external historical data to obtain a first demand set, a demand analysis is performed on an internal historical data set to obtain a second demand set, and the first demand set and the second demand set are combined to obtain a demand data set.

[0036] In this embodiment, external historical data refers to data collected from the market and the environment, such as market price data: records of market price fluctuations of coal; competitor analysis data: prices and supply conditions of other coal suppliers.

[0037] In this embodiment, internal historical data refers to data collected within the enterprise, such as sales records: coal sales volume and customer information within the company; inventory records: coal inventory status recorded in the company's inventory management system; production records: coal mine production efficiency and equipment maintenance records.

[0038] In this embodiment, the first supply set is a set obtained after supply analysis based on external historical data, for example, external suppliers: the supply capacity and price of a large coal mine, a local small coal mine, etc.; the second supply set is a set obtained after supply analysis based on internal historical data, for example, internal suppliers: the production capacity and inventory status of a company's own coal mines.

[0039] In this embodiment, the first demand set is a set obtained by demand analysis based on external historical data, for example, market demand data: such as the demand for coal in a certain area and its changing trend; the second demand set is a set obtained by demand analysis based on internal historical data, for example, internal customer demand: such as the demand records for coal by various departments within the company.

[0040] The working principle and beneficial effects of the above technical solution are: by determining the goal of fuel optimization, classifying and collecting historical data of the coal supply chain into external and internal data. Supply and demand analysis is performed on external data and internal data respectively, and multiple supply and demand sets are obtained, which are finally integrated to form supply data sets and demand data sets, providing a basis for accurate supply and demand forecasting, improving the accuracy and reliability of supply and demand forecasting, enhancing the flexibility of scheduling, and reducing information lag.

[0041] Embodiment 3: The embodiment of the present invention provides a fuel optimization scheduling method based on coal supply chain prediction, extracting supply features from the supply data set, extracting demand features from the demand data set, fusing the supply features with the demand features to form a comprehensive feature set, and constructing a prediction model for the supply and demand relationship of the coal supply chain based on the comprehensive feature set, including: Performing data structure analysis on the supply data set to determine a first key field of the supply data set, extracting the first key field as a first supply feature, performing feature derivation based on the first supply feature to obtain a second supply feature, and combining the first supply feature and the second supply feature to determine a supply feature set; Performing data structure analysis on the demand data set to determine a second key field of the demand data set, extracting the second key field as a first demand feature, performing feature derivation based on the first demand feature to obtain a second demand feature, and combining the first demand feature and the second demand feature to determine a demand feature set; A logical analysis is performed on the goals to be achieved in fuel optimization to derive the target logic, the demand characteristics and supply characteristics are combined based on the target logic to form a comprehensive feature set, and a prediction model for the supply and demand relationship of the coal supply chain is constructed based on the comprehensive feature set.

[0042] In this embodiment, data structure analysis is the understanding and evaluation of the composition, format and various fields of the data set, with the purpose of identifying the relationship between key fields and data, for example, analyzing the supply data set to determine fields such as "supplier ID", "supply quantity", "price", "delivery time", etc.

[0043] In this embodiment, the first key field refers to a field that can uniquely identify a record in the supply data set, such as supplier ID: a unique identifier for each supplier; the first supply feature is a feature extracted from the first key field for subsequent analysis, such as supplier type: such as "local supplier" or "external supplier"; the second supply feature is a new feature derived from the first supply feature, such as supply stability: a measure of the degree of fluctuation in supply volume.

[0044] In this embodiment, the feature derivation process refers to the process of generating new features based on existing features, for example, deriving the "supply stability" feature from the "supply quantity" field and calculating the standard deviation of the supply quantity in the past few months.

[0045] In this embodiment, the second key field refers to a field that can uniquely identify a record in the demand data set, such as customer ID: a unique identifier for each customer; the first demand feature is a feature extracted from the second key field, such as customer type: such as "industrial customer" or "retail customer"; the second demand feature is a new feature derived from the first demand feature, such as demand volatility: a measure of the degree of change in customer demand; the demand feature set is a feature set that combines the first demand feature with the second demand feature, such as a feature set containing "customer type" and "demand volatility".

[0046] In this embodiment, the target logic refers to the analysis of logical relationships and priorities among the goals to be achieved in fuel optimization, such as the trade-off between cost minimization and service level improvement.

[0047] In this embodiment, the process of logical analysis is to systematically analyze the goals to be achieved to determine the relationship and impact between the goals, for example, analyzing how to reduce costs by optimizing supply chain management without affecting delivery time.

[0048] In this embodiment, the combination process is to combine the demand characteristics with the supply characteristics to form a comprehensive feature set, for example, combining "customer type" and "supplier type" to form a "customer and supplier matching" feature.

[0049] The working principle and beneficial effects of the above technical solution are: through structural analysis of the supply and demand data sets of the coal supply chain, identifying key fields and deriving features, forming a comprehensive feature set, based on target logic, combining demand and supply characteristics, building a prediction model to achieve accurate supply and demand forecasting and flexible scheduling, and improve supply chain efficiency.

[0050] Embodiment 4: The embodiment of the present invention provides a fuel optimization scheduling method based on coal supply chain prediction, which constructs a scheduling model to output a fuel scheduling plan according to the prediction result output by the prediction model, and performs fuel scheduling, including: Extract the forecast results of fuel demand from the forecast model, identify the relevant factors affecting the scheduling process based on the goals to be achieved in fuel optimization, determine the constraints in the scheduling process based on the relevant factors, and select the optimization algorithm to build the scheduling model according to the forecast results and the constraints; A fuel scheduling plan is output according to the scheduling model, and the fuel scheduling plan is executed to obtain an actual scheduling result.

[0051] In this embodiment, relevant factors include external and internal factors that affect scheduling, such as changes in market demand, supply capacity, transportation capacity, inventory levels, weather factors, etc. For example, seasonal changes in market demand (such as increased demand for heating in winter) and transportation capacity (such as the number of transport vehicles that can be dispatched each day).

[0052] In this embodiment, the constraints are restrictions that must be followed during the scheduling process, such as maximum inventory capacity, minimum delivery quantity, transportation time limit, supplier delivery capability, etc. For example, the maximum inventory capacity (such as a warehouse can store up to 500 tons of fuel).

[0053] The working principle and beneficial effects of the above technical solution are: extracting fuel demand forecast results from the forecast model, identifying relevant factors affecting scheduling and determining constraints, selecting an optimization algorithm based on these constraints to build a scheduling model, outputting a fuel scheduling plan and executing it, and finally obtaining the actual scheduling results to achieve efficient fuel scheduling and resource allocation, and based on real-time data and forecast results, enhancing the scientificity and accuracy of scheduling decisions.

[0054] Embodiment 5: The embodiment of the present invention provides a fuel optimization scheduling method based on coal supply chain prediction, selecting an optimization algorithm according to the prediction results and constraint conditions, and constructing a scheduling model, including: Determine the objective function based on the target to be achieved in fuel optimization, randomly generate initial solutions according to the constraint conditions and the prediction results, calculate the objective function value of each initial solution, and set a dynamic threshold to make an excellent judgment on the objective function value of each initial solution to obtain an excellent solution; Extracting the excellent features of the excellent solution to set the disturbance range, and performing local search for the excellent solution based on the disturbance range to generate multiple candidate solutions, calculating the objective function values ​​of the candidate solutions, and judging the objective function values ​​of the candidate solutions according to the dynamic threshold to obtain the final solution; A scheduling model is constructed based on the final solution.

[0055] In this embodiment, the objective function is a performance indicator of a quantitative scheduling solution, which generally includes cost, efficiency, supply and demand balance, etc. For example, minimizing total transportation costs, maximizing supply chain response speed, etc.

[0056] In this embodiment, the initial solution is a scheduling scheme randomly generated during the optimization process for subsequent evaluation and improvement, such as a scheme for randomly allocating fuel supply to different demand points.

[0057] In this embodiment, the process of excellent determination is to compare the objective function value of the initial solution with the dynamic threshold. If the objective function value is lower than the threshold, it is determined to be an excellent solution.

[0058] In this embodiment, the process of setting the disturbance range is to analyze the characteristics of the excellent solution (such as supply and demand) and determine the appropriate upper and lower floating range to maintain the feasibility of the solution; the disturbance range is a variable range set based on the excellent solution and is used to generate new candidate solutions, for example, a fluctuation of 10% above and below a specific fuel distribution amount.

[0059] In this embodiment, local search is a process of performing detailed exploration near an excellent solution to find a better solution, making small adjustments to the excellent solution within the disturbance range, calculating the objective function value of the adjusted new solution, and judging its quality.

[0060] In this embodiment, candidate solutions are multiple new scheduling schemes generated by local search, which are awaiting further evaluation. For example, different fuel allocation schemes are formed after adjustment based on excellent solutions. After the final solution is evaluated and screened, the best scheduling scheme is determined. The objective function values ​​of all candidate solutions are calculated, and the best performing solution is selected based on the dynamic threshold.

[0061] In this embodiment, the dynamic threshold is a criterion that is dynamically adjusted based on current market conditions, historical data, and forecast results. For example, during peak demand periods, the threshold may be stricter; when supply is sufficient, the threshold may be relaxed.

[0062] The working principle and beneficial effects of the above technical solution are: by setting the objective function and constraints, randomly generating the initial solution and calculating its objective function value, using the dynamic threshold to screen out the excellent solution, performing local search based on the characteristics of the excellent solution, generating candidate solutions and evaluating their performance, and finally obtaining the final solution, and finally building an efficient scheduling model to improve the accuracy and flexibility of fuel scheduling. Embodiment 6: The embodiment of the present invention provides a fuel optimization scheduling method based on coal supply chain prediction, and sets a dynamic threshold to perform excellent judgment on the objective function value, including: ,in, Indicates at time t Dynamic threshold of Indicates a point in history The objective function value function of Indicates the current time point t The objective function value function of Indicates the current time point t Distribution function of the objective function value; Indicates the current time point t The dynamic threshold at the previous time point; Indicates a point in history; t Indicates the current time point; Indicates time period Neidii1 The objective function value of the initial solution; Indicates time period Neidi i2 The objective function value of the initial solution; k1 Indicates from The total number of time steps back through time; k2 Indicates from t The total number of time steps back through time; j1 Indicates from Going back in time j1 step; j2 Indicates from t Going back in time j2 step; Indicates time point The number of initial solutions at time t; Indicates time point The number of initial solutions at time t; Represents the adjustment factor of the mean change of the objective function value; Indicates the influence factor of the objective function value at the historical time point on the objective function value at the current time point; Represents the rate of change factor of the objective function value; Represents the distribution function change adjustment factor; Indicates the acceleration adjustment factor of the distribution function change; When the objective function value of the initial solution at the same time point is higher than the dynamic threshold, the corresponding initial solution is an excellent solution.

[0063] The working principle and beneficial effects of the above technical solution are: by setting a dynamic threshold to determine the excellence of the initial solution, combining the historical objective function value and the current objective function value, dynamically adjusting the threshold to reflect changes in energy demand and supply. When the initial solution objective function value at the same time point is higher than the dynamic threshold, the solution is determined to be an excellent solution, thereby optimizing the scheduling strategy and improving the flexibility and responsiveness of the supply chain.

[0064] Embodiment 7: The embodiment of the present invention provides a fuel optimization scheduling method based on coal supply chain prediction, which compares the actual scheduling result with the prediction result to obtain a supply and demand comparison result, and optimizes the prediction model based on the supply and demand comparison result, including: Perform error analysis on the actual scheduling results and the predicted results to determine the error situation at each stage of the fuel scheduling plan, and obtain a supply and demand comparison result based on the error situation; Conduct regularity analysis on the supply and demand comparison results to determine potential regularities of supply and demand changes, and optimize the prediction model based on the potential regularities of supply and demand changes.

[0065] In this embodiment, error analysis is a process of systematically evaluating the difference between the actual scheduling results and the predicted results. For example, assuming that the predicted coal demand for a certain time period is 1,000 tons, and the actual demand is 1,200 tons, the error is 200 tons.

[0066] In this embodiment, the error situation refers to the degree of deviation between the actual demand and the predicted demand in different stages (such as day, week, month). For example, in one month, the actual demand in a certain week is 1,500 tons, while the predicted demand is 1,300 tons, and the error situation is +200 tons; while in another week, the actual demand is 900 tons, and the predicted demand is 1,000 tons, and the error situation is -100 tons.

[0067] In this embodiment, the pattern analysis is an in-depth study of the supply and demand comparison results to identify trends and patterns in supply and demand changes. For example, the analysis found that demand usually increases in winter, while supply decreases during the peak summer season. This seasonal change can be used as a pattern.

[0068] In this embodiment, the potential laws of supply and demand changes refer to the factors and trends that affect supply and demand fluctuations discovered through long-term observations. For example, historical data shows that economic growth rate is positively correlated with coal demand, and weather changes can also affect demand. For example, cold weather can lead to increased demand.

[0069] In this embodiment, based on the underlying patterns of supply and demand changes, the parameters or algorithms of the prediction model are adjusted to more accurately reflect future supply and demand conditions. For example, if it is found that cold weather increases coal demand, weather factors can be added as variables in the prediction model and trained using a machine learning algorithm to improve the accuracy of the prediction.

[0070] The working principle and beneficial effects of the above technical solution are: by conducting error analysis on the actual scheduling results and the predicted results, the supply and demand differences in each stage are identified, and based on these errors, the potential laws of supply and demand changes are further analyzed, and the prediction model is optimized accordingly, thereby improving the accuracy and flexibility of future scheduling and adapting to the ever-changing market demands in the coal supply chain.

[0071] Embodiment 8: The embodiment of the present invention provides a fuel optimization scheduling method based on coal supply chain forecasting, which performs regularity analysis on the supply and demand comparison result, determines the potential regularity of supply and demand changes, and optimizes the forecasting model based on the potential regularity of supply and demand changes, including: Identify abnormalities in the supply and demand comparison results, determine abnormal points, and trace the causes of the abnormal points. Based on the abnormal points and causes, summarize the potential patterns of supply and demand changes; The potential rules and anomalies are incorporated into the prediction model to optimize the prediction model.

[0072] In this embodiment, an abnormal point refers to a situation in which the supply and demand comparison results significantly deviate from the normal range or expected value. These points usually manifest as a significant difference between actual demand or supply and the predicted value. For example, if the actual coal demand in a certain week is 2,000 tons, but the forecast is only 1,200 tons, then the demand data for that week can be regarded as an abnormal point because its deviation exceeds the normal fluctuation range.

[0073] In this embodiment, cause tracing is an in-depth analysis of the identified anomalies to determine the specific factors or causes that lead to the anomalies, which usually involves an investigation of relevant data and background information. For example, after identifying the demand anomaly in a certain week, factors such as weather data, economic activities, and policy changes may be analyzed to find that the abnormal demand was due to the extremely cold weather that week, which led to a surge in coal demand. Through this analysis, the specific causes can be traced back, thereby providing a reference for subsequent forecasts and scheduling.

[0074] The working principle and beneficial effects of the above technical solution are: by identifying anomalies in the supply and demand comparison results, determining the abnormal points and tracing their causes, summarizing the potential laws of supply and demand changes, incorporating these laws and anomalies into the prediction model, and optimizing the model parameters to improve the accuracy and responsiveness of future supply and demand forecasts, thereby enhancing the flexibility of the coal supply chain, incorporating potential laws into the model, and improving the accuracy and reliability of the forecast.

[0075] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit it. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or make equivalent replacements for some of the technical features therein. However, these modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A fuel optimization scheduling method based on coal supply chain forecasting, characterized in that: include: Step 1: Based on the goals to be achieved in fuel optimization, relevant historical data on the coal supply chain are collected, and supply analysis and demand analysis are performed on the relevant historical data of the coal supply chain to obtain a supply data set and a demand data set; Step 2: extracting supply features from the supply data set, extracting demand features from the demand data set, fusing the supply features with the demand features to form a comprehensive feature set, and constructing a prediction model for the supply and demand relationship of the coal supply chain based on the comprehensive feature set; Step 3: Based on the prediction results output by the prediction model, a scheduling model is constructed, a fuel scheduling plan is output, fuel scheduling is performed, and actual scheduling results are obtained; Step 4: Compare the actual scheduling results with the forecast results to obtain the supply and demand comparison results, and optimize the forecast model based on the supply and demand comparison results.

2. A fuel optimization scheduling method based on coal supply chain forecasting according to claim 1, characterized in that: Based on the goals to be achieved in fuel optimization, relevant historical data on the coal supply chain are collected, and supply analysis and demand analysis are performed on the relevant historical data of the coal supply chain to obtain supply data sets and demand data sets, including: Obtaining a target to be achieved for fuel optimization, collecting relevant historical data of the coal supply chain based on the target to be achieved for fuel optimization, and classifying the relevant historical data of the coal supply chain to obtain external historical data and internal historical data; Performing supply analysis on external historical data to obtain a first supply set, performing supply analysis on internal historical data to obtain a second supply set, and combining the first supply set and the second supply set to obtain a supply data set; A demand analysis is performed on external historical data to obtain a first demand set, a demand analysis is performed on an internal historical data set to obtain a second demand set, and the first demand set and the second demand set are combined to obtain a demand data set.

3. The method for optimizing fuel scheduling based on coal supply chain forecasting according to claim 1 is characterized in that: Extracting supply features from the supply data set, extracting demand features from the demand data set, fusing the supply features with the demand features to form a comprehensive feature set, and constructing a prediction model for the supply and demand relationship of the coal supply chain based on the comprehensive feature set, including: Performing data structure analysis on the supply data set to determine a first key field of the supply data set, extracting the first key field as a first supply feature, performing feature derivation based on the first supply feature to obtain a second supply feature, and combining the first supply feature and the second supply feature to determine a supply feature set; Performing data structure analysis on the demand data set to determine a second key field of the demand data set, extracting the second key field as a first demand feature, performing feature derivation based on the first demand feature to obtain a second demand feature, and combining the first demand feature and the second demand feature to determine a demand feature set; A logical analysis is performed on the goals to be achieved in fuel optimization to derive the target logic, the demand characteristics and supply characteristics are combined based on the target logic to form a comprehensive feature set, and a prediction model for the supply and demand relationship of the coal supply chain is constructed based on the comprehensive feature set.

4. The method for optimizing fuel scheduling based on coal supply chain forecasting according to claim 1, characterized in that: According to the prediction results output by the prediction model, a scheduling model is constructed to output the fuel scheduling plan and perform fuel scheduling, including: Extract the forecast results of fuel demand from the forecast model, identify the relevant factors affecting the scheduling process based on the goals to be achieved in fuel optimization, determine the constraints in the scheduling process based on the relevant factors, and select the optimization algorithm to build the scheduling model according to the forecast results and the constraints; A fuel scheduling plan is output according to the scheduling model, and the fuel scheduling plan is executed to obtain an actual scheduling result.

5. A fuel optimization scheduling method based on coal supply chain forecasting according to claim 4, characterized in that: An optimization algorithm is selected according to the prediction results and constraints, and a scheduling model is constructed, including: Determine the objective function based on the target to be achieved in fuel optimization, randomly generate initial solutions according to the constraint conditions and the prediction results, calculate the objective function value of each initial solution, and set a dynamic threshold to make an excellent judgment on the objective function value of each initial solution to obtain an excellent solution; Extracting the excellent features of the excellent solution to set the disturbance range, and performing local search for the excellent solution based on the disturbance range to generate multiple candidate solutions, calculating the objective function values ​​of the candidate solutions, and judging the objective function values ​​of the candidate solutions according to the dynamic threshold to obtain the final solution; A scheduling model is constructed based on the final solution.

6. A fuel optimization scheduling method based on coal supply chain forecasting according to claim 5, characterized in that: And set a dynamic threshold to make an excellent judgment on the objective function value, including: ,in, Indicates at time t Dynamic threshold of Indicates a point in history The objective function value function of Indicates the current time point t The objective function value function of Indicates the current time point t Distribution function of the objective function value; Indicates the current time point t The dynamic threshold at the previous time point; Indicates a point in history; t Indicates the current time point; Indicates time period Neidi i1 The objective function value of the initial solution; Indicates time period Neidi i2 The objective function value of the initial solution; k1 Indicates from The total number of time steps back and forward; k2 Indicates from t The total number of time steps back and forward; j1 Indicates from Going back in time j1 step; j2 Indicates from t Going back in time j2 step; Indicates time point The number of initial solutions at time t; Indicates time point The number of initial solutions at time t; Represents the adjustment factor of the mean change of the objective function value; Indicates the influence factor of the objective function value at the historical time point on the objective function value at the current time point; Represents the rate of change factor of the objective function value; Represents the distribution function change adjustment factor; Indicates the acceleration adjustment factor of the distribution function change; When the objective function value of the initial solution at the same time point is higher than the dynamic threshold, the corresponding initial solution is an excellent solution.

7. The method for optimizing fuel scheduling based on coal supply chain forecasting according to claim 1, characterized in that: Compare the actual scheduling results with the forecast results to obtain the supply and demand comparison results, and optimize the forecast model based on the supply and demand comparison results, including: Perform error analysis on the actual scheduling results and the predicted results to determine the error situation at each stage of the fuel scheduling plan, and obtain a supply and demand comparison result based on the error situation; Conduct regularity analysis on the supply and demand comparison results to determine potential regularities of supply and demand changes, and optimize the prediction model based on the potential regularities of supply and demand changes.

8. The method for optimizing fuel scheduling based on coal supply chain forecasting according to claim 1, characterized in that: Performing regularity analysis on the supply and demand comparison results to determine potential regularities of supply and demand changes, and optimizing the prediction model based on the potential regularities of supply and demand changes, including: Identify abnormalities in the supply and demand comparison results, determine abnormal points, and trace the causes of the abnormal points. Based on the abnormal points and causes, summarize the potential patterns of supply and demand changes; The potential rules and anomalies are incorporated into the prediction model to optimize the prediction model.

Images