All-digital fuel cost analysis and control platform

Abstract

The invention discloses an all-digital fuel cost analysis and control platform which comprises a fuel information module, a fuel prediction module, a fuel regulation and control module and a fuel accounting module. The fuel information module is used for storing historical fuel information and acquiring real-time fuel information; the fuel prediction module is used for predicting the next-period fuel price and the next-period fuel supply information, and analyzing the abnormal fuel cost information by using an abnormal model to obtain an abnormal reason; the fuel regulation and control module is used for obtaining a next-stage fuel strategy according to the next-stage fuel price and the next-stage fuel supply information and dynamically adjusting the fuel strategy; the fuel accounting module is used for accounting the fuel cost in unit time and performing transaction risk level evaluation on each transaction client according to historical settlement data; the fuel strategy is formulated by predicting the fuel price and the fuel demand in the future, the problems of fuel insufficiency and fuel excess are avoided, the fuel cost is saved, and the fuel benefit is obtained.

Description

Technical Field

[0001] The present invention relates to the technical field of data analysis, and in particular to a fully digital fuel cost analysis and control platform. Background Art

[0002] In recent years, with the continuous changes in the power coal market environment and the continuous advancement of my country's coal-fired power enterprise policies, on the one hand, the spot transactions carried out in the power market have gradually entered a mature operating state, and on the other hand, as the competition in the power market is fierce, the transaction prices of bulk commodities such as coal continue to break historical records, and the cost pressure faced by thermal power generation companies is increasing. Therefore, cost control of power generation companies has become crucial at this stage.

[0003] At present, the Chinese invention patent with application number 202210210825.0 provides a smart fuel management and control platform. Although it uses the information brought by automation, informatization and intelligence to comprehensively and comprehensively reduce the cost of coal entering the furnace, improve the economic benefits of the enterprise, and intelligently control coal burning, it does not make detailed predictions and assessments of future trends in fuel prices and fuel demand, which is not conducive to the formulation of fuel procurement strategies, and does not conduct transaction risk assessments for trading customers, and does not guarantee the security of fuel transactions. Summary of the invention

[0004] The technical problem solved by the present invention is that the prior art does not make detailed predictions and assessments on the future trends of fuel prices and fuel demand, which is not conducive to the formulation of fuel procurement strategies, and does not conduct transaction risk assessments for trading customers, thereby failing to ensure the security of fuel transactions.

[0005] In order to solve the above technical problems, the present invention provides a fully digital fuel cost analysis and control platform, including a fuel information module, a fuel prediction module, a fuel control module and a fuel accounting module; The fuel information module is used to store historical fuel information and obtain real-time fuel information; The fuel prediction module is used to obtain the next period fuel price and the next period fuel supply information based on the historical fuel information by using correlation analysis, and obtain the abnormal fuel cost information from the fuel accounting module, and use the abnormal model to analyze the abnormal fuel cost information to obtain the abnormal cause; The fuel control module is used to obtain the next-period fuel strategy according to the next-period fuel price and the next-period fuel supply information, and adjust the fuel strategy according to abnormal conditions in the real-time fuel information; The fuel accounting module is used to calculate the fuel cost per unit time, obtain abnormal fuel cost information according to the expected threshold, and conduct transaction risk level assessment for each transaction customer based on historical settlement data; As a preferred solution of the all-digital fuel cost analysis and control platform described in the present invention, wherein: The fuel information module includes a historical fuel unit and a real-time fuel unit; The historical fuel unit is used to obtain and store historical fuel information, wherein the historical fuel information includes historical fuel purchase information, historical fuel transportation information, historical fuel storage information and historical fuel demand information, and the historical fuel purchase information includes historical fuel types, historical fuel prices and historical fuel transaction records; The real-time fuel unit is used to obtain fuel purchase information through the fuel purchase terminal, obtain fuel transportation information through the fuel transportation terminal, obtain fuel storage information through the fuel storage terminal, and obtain fuel demand information through the fuel sales terminal; As a preferred solution of the all-digital fuel cost analysis and control platform described in the present invention, wherein: The fuel prediction module includes a fuel prediction unit and a fuel anomaly unit; The fuel prediction unit is used to predict the next period of fuel price based on the historical fuel price by using correlation analysis, and predict the next period of fuel demand information by using correlation analysis based on the historical fuel demand information; The fuel anomaly unit is used to obtain historical fuel anomaly information and abnormal fuel cost information, establish an anomaly model based on the historical fuel anomaly information, use the anomaly model to obtain the cause of the anomaly based on the abnormal fuel cost information, and adjust the correlation coefficient of the influencing factor used to predict the next period of fuel price and the next period of fuel demand information based on the abnormal cause; As a preferred solution of the all-digital fuel cost analysis and control platform described in the present invention, wherein: The fuel control module includes a fuel strategy unit and a fuel control unit; The fuel strategy unit is used to formulate a fuel strategy based on the next period's fuel price and the next period's fuel supply and demand information; The fuel control unit is used to dynamically adjust the fuel strategy according to real-time fuel information, and the fuel strategy includes fuel procurement strategy, fuel transportation strategy and fuel storage strategy; As a preferred solution of the all-digital fuel cost analysis and control platform described in the present invention, wherein: The fuel accounting module includes a historical settlement unit and a fuel settlement unit; The historical settlement unit is used to store historical settlement data, and analyze different transaction customers according to the historical settlement data, obtain transaction customer feature portraits, and evaluate transaction risk levels according to the transaction customer feature portraits; The fuel settlement unit is used to calculate the fuel cost per unit time, and obtain fuel cost information when the fuel cost per unit time exceeds an expected threshold, define it as abnormal fuel cost information, and send the abnormal fuel cost information to the fuel abnormality unit to obtain the abnormality cause; As a preferred solution of the all-digital fuel cost analysis and control platform described in the present invention, wherein: The fuel forecasting unit includes a fuel price forecasting subunit and a fuel demand forecasting subunit; The fuel price prediction subunit is used to obtain the influencing factors that may affect the fuel price, and use the Pearson product-moment correlation coefficient formula to obtain the correlation coefficient of each fuel price influencing factor. If the correlation coefficient of the influencing factor is greater than 0.5, it is defined as the fuel price influencing factor, and the next period of fuel price is obtained according to the correlation coefficient of the fuel price influencing factor; The fuel demand prediction subunit is used to obtain the influencing factors that may affect the fuel demand information, and use the Pearson product-moment correlation coefficient formula to obtain the correlation coefficient of the influencing factors of each fuel demand information. If the correlation coefficient of the influencing factor of the fuel demand information is greater than 0.5, it is defined as the fuel demand influencing factor, and the next period of fuel demand information is obtained according to the correlation coefficient of the fuel demand influencing factor; As a preferred solution of the all-digital fuel cost analysis and control platform described in the present invention, wherein: The establishing of an abnormal model according to the historical fuel abnormal information comprises: Obtain historical fuel abnormality information, perform data preprocessing on the historical fuel abnormality information, perform data annotation on the historical fuel abnormality information after data preprocessing, and obtain the abnormality cause, input the historical fuel abnormality information and the abnormality cause into the machine learning model for training, and use the historical fuel abnormality information as input and the abnormality cause as output to establish an abnormality model, until the model fit is greater than or equal to 0.9, then the training ends; As a preferred solution of the all-digital fuel cost analysis and control platform described in the present invention, wherein: The historical settlement unit includes a risk model subunit and a risk assessment subunit; The risk model subunit is used to obtain historical settlement data, perform data preprocessing and feature extraction on the historical settlement data, classify the historical settlement data according to data features, and define primary classification labels, perform in-depth feature mining on different categories of historical settlement data, define secondary classification labels and third-class classification labels respectively, and define transaction risk levels of different transaction customers according to the superposition of classification labels; The risk model subunit is used to divide the historical settlement data and transaction risk levels into a training set, a test set and a validation set, train the machine learning model using the training set, and validate and optimize the machine learning model using the test set and the validation set; The risk assessment subunit is used to obtain historical settlement data of transaction customers, and input the historical settlement data into the machine learning model, obtain and update the risk assessment level of each transaction customer, and set corresponding fuel transaction caps and contractable price ranges for transaction customers according to different risk assessment levels; As a preferred solution of the all-digital fuel cost analysis and control platform described in the present invention, wherein: The fuel control unit is used to dynamically adjust the fuel strategy according to the real-time fuel information, including: Acquiring real-time fuel information, the real-time fuel information including fuel purchase information, fuel transportation information, fuel storage information and fuel demand information; Obtaining a fuel price according to the fuel purchase information, obtaining a fuel inventory according to the fuel storage information, and obtaining a fuel demand according to the fuel demand information; If the fuel price, fuel inventory or fuel demand exceeds the threshold range, it is sent to the fuel strategy unit, the corresponding value of the benefit function in the fuel strategy unit is updated, and the fuel strategy is re-obtained; As a preferred solution of the all-digital fuel cost analysis and control platform described in the present invention, wherein: The fuel strategy formulated according to the next period fuel price and the next period fuel supply and demand information includes: Obtain historical fuel procurement information, historical fuel transportation information, historical fuel storage information, historical fuel demand information and historical fuel benefit information, obtain a benefit function and a benefit evaluation function based on the historical fuel procurement information, historical fuel transportation information, historical fuel storage information, historical fuel demand information and historical fuel benefit information, input the next period's fuel price and the next period's fuel supply and demand information into the benefit function, continuously and randomly adjust the various fuel quantities in the benefit function, and use the benefit evaluation function to evaluate the fuel benefits generated by each randomly generated fuel strategy until a fuel strategy with the best evaluation result is obtained, and this fuel strategy is used as the final fuel strategy.

[0006] Beneficial effects of the present invention: The present invention predicts future fuel prices and fuel demands through historical fuel information, and formulates fuel strategies based on fuel prices and fuel demands, thereby avoiding the problems of fuel shortage and fuel surplus, ensuring the appropriate purchase of fuel, and being conducive to saving fuel costs and obtaining fuel benefits.

[0007] An abnormal model is established based on historical fuel abnormality information, and abnormal fuel cost information is input into the abnormal model to obtain the abnormal cause, which is conducive to avoiding the occurrence of abnormal fuel cost information, effectively controlling the input of fuel cost and saving fuel cost.

[0008] Based on historical fuel settlement data, risk level assessment is conducted for each trading customer, which avoids fuel quality problems caused by trading risks, ensures the security of fuel transactions, and saves fuel costs to a certain extent. BRIEF DESCRIPTION OF THE DRAWINGS

[0009] Figure 1 A basic flow chart of a fully digital fuel cost analysis and control platform provided for one embodiment of the present invention. DETAILED DESCRIPTION

[0010] In order to make the above-mentioned objects, features and advantages of the present invention more obvious and easy to understand, the specific implementation methods of the present invention are described in detail below in conjunction with the drawings. It is obvious that the described embodiments are only part of the embodiments of the present invention, but not all of the embodiments.

[0011] Example, see Figure 1 , which is an embodiment of the present invention, provides a fully digital fuel cost analysis and control platform, including a fuel information module, a fuel prediction module, a fuel control module and a fuel accounting module; The fuel information module is used to store historical fuel information and obtain real-time fuel information; The fuel prediction module is used to obtain the next period fuel price and the next period fuel supply information based on the historical fuel information by using correlation analysis, and obtain the abnormal fuel cost information from the fuel accounting module, and use the abnormal model to analyze the abnormal fuel cost information to obtain the abnormal cause; The fuel control module is used to obtain the next-period fuel strategy according to the next-period fuel price and the next-period fuel supply information, and adjust the fuel strategy according to abnormal conditions in the real-time fuel information; The fuel accounting module is used to calculate the fuel cost per unit time, obtain abnormal fuel cost information according to the expected threshold, and evaluate the transaction risk level for each transaction customer based on historical settlement data.

[0012] In this embodiment, future fuel prices and fuel demands are predicted through historical fuel information, and a fuel strategy is formulated based on the fuel price and fuel demand, thereby avoiding the problems of fuel shortage and fuel surplus, ensuring the appropriate purchase of fuel, and helping to save fuel costs and obtain fuel benefits.

[0013] An abnormal model is established based on historical fuel abnormality information, and abnormal fuel cost information is input into the abnormal model to obtain the abnormal cause, which is conducive to avoiding the occurrence of abnormal fuel cost information, effectively controlling the input of fuel cost and saving fuel cost.

[0014] Based on historical fuel settlement data, risk level assessment is conducted for each trading customer, which avoids fuel quality problems caused by trading risks, ensures the security of fuel transactions, and saves fuel costs to a certain extent.

[0015] The fuel information module includes a historical fuel unit and a real-time fuel unit; The historical fuel unit is used to obtain and store historical fuel information, wherein the historical fuel information includes historical fuel purchase information, historical fuel transportation information, historical fuel storage information and historical fuel demand information, and the historical fuel purchase information includes historical fuel types, historical fuel prices and historical fuel transaction records; The real-time fuel unit is used to obtain fuel purchase information through a fuel purchase terminal, obtain fuel transportation information through a fuel transportation terminal, obtain fuel storage information through a fuel storage terminal, and obtain fuel demand information through a fuel selling terminal.

[0016] The historical fuel purchase information includes historical fuel types, historical fuel prices and historical fuel transaction records; The historical fuel transportation information includes historical fuel transportation equipment, historical fuel transportation routes, historical fuel types, historical fuel loading, historical fuel transportation evaporation and historical fuel transportation costs; The historical fuel storage information includes historical fuel storage location, historical fuel storage mode, historical fuel type, historical fuel storage amount and historical fuel storage evaporation amount; The historical demand information includes historical fuel demanders and historical fuel demand amounts.

[0017] The fuel purchase information includes fuel purchase price and fuel transaction record; The fuel transportation information includes fuel transportation equipment, fuel transportation routes, fuel types, fuel loading, fuel transportation evaporation and fuel transportation costs; The fuel storage information includes fuel storage location, fuel storage mode, fuel type, fuel storage amount and fuel storage evaporation amount; The fuel demand information includes the fuel demander and the fuel demand amount.

[0018] In this embodiment, obtaining historical fuel information and real-time fuel information provides detailed and specific data support for predicting the next period's fuel price, the next period's fuel demand information, and adjusting the fuel strategy in real time.

[0019] The fuel prediction module includes a fuel prediction unit and a fuel anomaly unit; The fuel prediction unit is used to predict the next period of fuel price based on the historical fuel price by using correlation analysis, and predict the next period of fuel demand information by using correlation analysis based on the historical fuel demand information; The fuel anomaly unit is used to obtain historical fuel anomaly information and abnormal fuel cost information, establish an anomaly model based on the historical fuel anomaly information, use the anomaly model to obtain the cause of the anomaly based on the abnormal fuel cost information, and adjust the correlation coefficient of the influencing factor used to predict the next period's fuel price and the next period's fuel demand information based on the abnormal cause.

[0020] The correlation analysis refers to analyzing two or more correlated variable elements to measure the degree of correlation between the two variable factors; The correlation analysis includes: The correlation coefficient between variables was measured using the Pearson product-moment correlation coefficient formula; Calculate correlation coefficients and perform hypothesis testing; Perform linear regression analysis to obtain a linear regression equation for actual value prediction; The correlation coefficient is a value that measures the degree of correlation between variables.

[0021] In this embodiment, correlation analysis is used to obtain the next period's fuel price and the next period's fuel demand information based on historical fuel information, and an abnormal model is established to obtain the abnormal reasons for the abnormal fuel cost information, providing an accurate data reference for obtaining fuel strategies. The acquisition of abnormal reasons is also conducive to the effective control of fuel costs.

[0022] The fuel control module includes a fuel strategy unit and a fuel control unit; The fuel strategy unit is used to formulate a fuel strategy based on the next period's fuel price and the next period's fuel supply and demand information; The fuel control unit is used to dynamically adjust the fuel strategy according to real-time fuel information, and the fuel strategy includes a fuel procurement strategy, a fuel transportation strategy and a fuel storage strategy.

[0023] The fuel purchasing strategy indicates the types of fuel purchased and the corresponding purchase quantities; The fuel transportation strategy represents the fuel transportation route, fuel transportation method and fuel transportation amount selected for each category of fuel; The fuel storage strategy indicates a fuel storage location, a fuel storage method, and a fuel storage amount selected for each type of fuel.

[0024] In this embodiment, the fuel strategy is obtained based on the next period's fuel price and the next period's fuel supply and demand information, and the fuel strategy is dynamically adjusted based on real-time fuel information, effectively avoiding abnormal situations of fuel shortage and fuel surplus, and ensuring appropriate investment in fuel costs and acquisition of fuel benefits.

[0025] The fuel accounting module includes a historical settlement unit and a fuel settlement unit; The historical settlement unit is used to store historical settlement data, and analyze different transaction customers according to the historical settlement data, obtain transaction customer feature portraits, and evaluate transaction risk levels according to the transaction customer feature portraits; The fuel settlement unit is used to calculate the fuel cost per unit time, and obtain fuel cost information when the fuel cost per unit time exceeds an expected threshold, define it as abnormal fuel cost information, and send the abnormal fuel cost information to the fuel abnormality unit to obtain the abnormality cause.

[0026] The feature profiling refers to the process of analyzing and mining the multi-dimensional information of historical settlement data to visualize the abstract information such as the characteristics, needs, and behaviors of transaction customers, and form a set of representative and recognizable labels; In this embodiment, each transaction customer is analyzed based on historical settlement data, and the transaction risk level is evaluated for different transaction customers, which provides specific data reference for ensuring fuel quality and transaction security, and saves fuel costs to a certain extent.

[0027] The fuel forecasting unit includes a fuel price forecasting subunit and a fuel demand forecasting subunit; The fuel price prediction subunit is used to obtain the influencing factors that may affect the fuel price, and use the Pearson product-moment correlation coefficient formula to obtain the correlation coefficient of each fuel price influencing factor. If the correlation coefficient of the influencing factor is greater than 0.5, it is defined as the fuel price influencing factor, and the next period of fuel price is obtained according to the correlation coefficient of the fuel price influencing factor; The fuel demand prediction subunit is used to obtain influencing factors that may affect the fuel demand information, and use the Pearson product-moment correlation coefficient formula to obtain the correlation coefficient of the influencing factors of each fuel demand information. If the correlation coefficient of the influencing factor of the fuel demand information is greater than 0.5, it is defined as the fuel demand influencing factor, and the next period of fuel demand information is obtained based on the correlation coefficient of the fuel demand influencing factor.

[0028] The Pearson product-moment correlation coefficient formula is expressed as follows: , Among them, r is the correlation coefficient, is the value of each target variable, is the mean of the target variable, is the value of the impact factor, is the mean of the impact factor, n is a natural number, and i is a natural number from 1 to n.

[0029] In this embodiment, correlation analysis is used to obtain the correlation coefficients of the factors affecting the fuel price and the correlation coefficients of the factors affecting the fuel demand information, which provides an effective basis for accurately predicting the next period of fuel price and the next period of fuel demand information.

[0030] The establishing of an abnormal model according to the historical fuel abnormal information comprises: Obtain historical fuel anomaly information, perform data preprocessing on the historical fuel anomaly information, perform data annotation on the historical fuel anomaly information after data preprocessing, and obtain the anomaly cause, input the historical fuel anomaly information and the anomaly cause into the machine learning model for training, and use the historical fuel anomaly information as input and the anomaly cause as output to establish an anomaly model, until the model fit is greater than or equal to 0.9, then the training ends.

[0031] Data annotation is the process of processing unprocessed voice, image, text, video and other data to transform them into machine-recognizable information. The machine learning model is trained by inputting massive training data to enable the model to grasp the data rules and then accurately classify or predict the newly input data; The model fit is used to evaluate the degree of agreement between the observed data and the expected values ​​of the model.

[0032] In this embodiment, machine learning model technology is used to establish an abnormal model based on historical fuel abnormality information, which provides an accurate model for obtaining abnormal causes for fuel abnormal cost information, ensures normal investment in fuel costs, and is conducive to saving fuel costs.

[0033] The historical settlement unit includes a risk model subunit and a risk assessment subunit; The risk model subunit is used to obtain historical settlement data, perform data preprocessing and feature extraction on the historical settlement data, classify the historical settlement data according to data features, and define primary classification labels, perform in-depth feature mining on different categories of historical settlement data, define secondary classification labels and third-class classification labels respectively, and define transaction risk levels of different transaction customers according to the superposition of classification labels; The risk model subunit is used to divide the historical settlement data and transaction risk levels into a training set, a test set and a validation set, train the machine learning model using the training set, and validate and optimize the machine learning model using the test set and the validation set; The risk assessment subunit is used to obtain the historical settlement data of trading customers, and input the historical settlement data into the machine learning model, obtain and update the risk assessment level of each trading customer, and set corresponding fuel trading upper limits and contractable price ranges for trading customers with different risk assessment levels.

[0034] The data preprocessing is to clean, convert, integrate, normalize and other processes on the original data before data analysis to improve the data quality; Feature extraction refers to extracting the most representative features from the original data so that the machine learning algorithm can better learn and predict the target; In this embodiment, the risk assessment level of each transaction customer is obtained by extracting features from historical settlement data, and the corresponding fuel transaction upper limit and contractable price range are set according to the risk assessment level, which effectively avoids transaction risks in fuel procurement, ensures fuel quality, and saves fuel costs to a certain extent.

[0035] The fuel control unit is used to dynamically adjust the fuel strategy according to the real-time fuel information, including: Acquiring real-time fuel information, the real-time fuel information including fuel purchase information, fuel transportation information, fuel storage information and fuel demand information; Obtaining a fuel price according to the fuel purchase information, obtaining a fuel inventory according to the fuel storage information, and obtaining a fuel demand according to the fuel demand information; If the fuel price, fuel inventory or fuel demand exceeds the threshold range, it is sent to the fuel strategy unit, the corresponding value of the benefit function in the fuel strategy unit is updated, and the fuel strategy is re-acquired.

[0036] In this embodiment, the fuel strategy is dynamically adjusted by acquiring real-time fuel information to ensure appropriate investment in fuel costs and save fuel costs.

[0037] The fuel strategy formulated according to the next period fuel price and the next period fuel supply and demand information includes: Obtain historical fuel procurement information, historical fuel transportation information, historical fuel storage information, historical fuel demand information and historical fuel benefit information, obtain a benefit function and a benefit evaluation function based on the historical fuel procurement information, historical fuel transportation information, historical fuel storage information, historical fuel demand information and historical fuel benefit information, input the next period's fuel price and the next period's fuel supply and demand information into the benefit function, continuously and randomly adjust the various fuel quantities in the benefit function, and use the benefit evaluation function to evaluate the fuel benefits generated by each randomly generated fuel strategy until a fuel strategy with the best evaluation result is obtained, and this fuel strategy is used as the final fuel strategy.

[0038] An abnormal model is established based on historical fuel abnormality information, and abnormal fuel cost information is input into the abnormal model to obtain the abnormal cause, which is conducive to avoiding the occurrence of abnormal fuel cost information, effectively controlling the input of fuel cost and saving fuel cost.

[0039] Based on historical fuel settlement data, risk level assessment is conducted for each trading customer, which avoids fuel quality problems caused by trading risks, ensures the security of fuel transactions, and saves fuel costs to a certain extent.

[0040] It should be understood by those skilled in the art that the embodiments of the present invention may be provided as methods, systems or computer program products. Therefore, the present invention may take the form of a complete hardware embodiment, a complete software embodiment or an embodiment combining software and hardware aspects. Moreover, the present invention may take the form of a computer program product implemented on one or more computer-usable storage media containing computer-usable program codes. Among them, the storage medium may be implemented by any type of volatile or non-volatile storage device or a combination thereof, such as static random access memory (Static Random Access Memory, referred to as SRAM), electrically erasable programmable read-only memory (Electrically Erasable Programmable Read-Only Memory, referred to as EEPROM), erasable programmable read-only memory (Erasable Programmable Read Only Memory, referred to as EPROM), programmable read-only memory (Programmable Red-Only Memory, referred to as PROM), read-only memory (Read-Only Memory, referred to as ROM), magnetic memory, flash memory, magnetic disk or optical disk. These computer program instructions may also be stored in a computer-readable memory capable of directing a computer or other programmable data processing device to operate in a specific manner, so that the instructions stored in the computer-readable memory produce an article of manufacture comprising an instruction device, which implements the process Figure 1 A process or multiple processes and / or boxes Figure 1 A function specified in one or more boxes.

[0041] 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 preferred embodiments, those skilled in the art should understand that the technical solutions of the present invention may be modified or replaced by equivalents without departing from the spirit and scope of the technical solutions of the present invention, which should all be included in the scope of the claims of the present invention.

Claims

1. A fully digital fuel cost analysis and control platform, characterized in that: It includes fuel information module, fuel prediction module, fuel control module and fuel accounting module; The fuel information module is used to store historical fuel information and obtain real-time fuel information; The fuel prediction module is used to obtain the next period fuel price and the next period fuel supply information based on the historical fuel information by using correlation analysis, and obtain the abnormal fuel cost information from the fuel accounting module, and use the abnormal model to analyze the abnormal fuel cost information to obtain the abnormal cause; The fuel control module is used to obtain the next-period fuel strategy according to the next-period fuel price and the next-period fuel supply information, and adjust the fuel strategy according to abnormal conditions in the real-time fuel information; The fuel accounting module is used to calculate the fuel cost per unit time, obtain abnormal fuel cost information according to the expected threshold, and evaluate the transaction risk level for each transaction customer based on historical settlement data.

2. A fully digital fuel cost analysis and control platform as claimed in claim 1, characterized in that: The fuel information module includes a historical fuel unit and a real-time fuel unit; The historical fuel unit is used to obtain and store historical fuel information, wherein the historical fuel information includes historical fuel purchase information, historical fuel transportation information, historical fuel storage information and historical fuel demand information, and the historical fuel purchase information includes historical fuel types, historical fuel prices and historical fuel transaction records; The real-time fuel unit is used to obtain fuel purchase information through a fuel purchase terminal, obtain fuel transportation information through a fuel transportation terminal, obtain fuel storage information through a fuel storage terminal, and obtain fuel demand information through a fuel selling terminal.

3. A fully digital fuel cost analysis and control platform as claimed in claim 1, characterized in that: The fuel prediction module includes a fuel prediction unit and a fuel anomaly unit; The fuel prediction unit is used to predict the next period of fuel price based on the historical fuel price by using correlation analysis, and predict the next period of fuel demand information by using correlation analysis based on the historical fuel demand information; The fuel anomaly unit is used to obtain historical fuel anomaly information and abnormal fuel cost information, establish an anomaly model based on the historical fuel anomaly information, use the anomaly model to obtain the cause of the anomaly based on the abnormal fuel cost information, and adjust the correlation coefficient of the influencing factor used to predict the next period's fuel price and the next period's fuel demand information based on the abnormal cause.

4. A fully digital fuel cost analysis and control platform as claimed in claim 1, characterized in that: The fuel control module includes a fuel strategy unit and a fuel control unit; The fuel strategy unit is used to formulate a fuel strategy based on the next period's fuel price and the next period's fuel supply and demand information; The fuel control unit is used to dynamically adjust the fuel strategy according to real-time fuel information, and the fuel strategy includes a fuel procurement strategy, a fuel transportation strategy and a fuel storage strategy.

5. A fully digital fuel cost analysis and control platform as claimed in claim 1, characterized in that: The fuel accounting module includes a historical settlement unit and a fuel settlement unit; The historical settlement unit is used to store historical settlement data, and analyze different transaction customers according to the historical settlement data, obtain transaction customer feature portraits, and evaluate transaction risk levels according to the transaction customer feature portraits; The fuel settlement unit is used to calculate the fuel cost per unit time, and obtain fuel cost information when the fuel cost per unit time exceeds an expected threshold, define it as abnormal fuel cost information, and send the abnormal fuel cost information to the fuel abnormality unit to obtain the abnormality cause.

6. A fully digital fuel cost analysis and control platform as claimed in claim 3, characterized in that: The fuel forecasting unit includes a fuel price forecasting subunit and a fuel demand forecasting subunit; The fuel price prediction subunit is used to obtain the influencing factors that may affect the fuel price, and use the Pearson product-moment correlation coefficient formula to obtain the correlation coefficient of each fuel price influencing factor. If the correlation coefficient of the influencing factor is greater than 0.5, it is defined as the fuel price influencing factor, and the next period of fuel price is obtained according to the correlation coefficient of the fuel price influencing factor; The fuel demand prediction subunit is used to obtain influencing factors that may affect the fuel demand information, and use the Pearson product-moment correlation coefficient formula to obtain the correlation coefficient of the influencing factors of each fuel demand information. If the correlation coefficient of the influencing factor of the fuel demand information is greater than 0.5, it is defined as the fuel demand influencing factor, and the next period of fuel demand information is obtained based on the correlation coefficient of the fuel demand influencing factor.

7. A fully digital fuel cost analysis and control platform as claimed in claim 3, characterized in that: The establishing of an abnormal model according to the historical fuel abnormal information comprises: Obtain historical fuel anomaly information, perform data preprocessing on the historical fuel anomaly information, perform data annotation on the historical fuel anomaly information after data preprocessing, and obtain the anomaly cause, input the historical fuel anomaly information and the anomaly cause into the machine learning model for training, and use the historical fuel anomaly information as input and the anomaly cause as output to establish an anomaly model, until the model fit is greater than or equal to 0.9, then the training ends.

8. A fully digital fuel cost analysis and control platform as claimed in claim 5, characterized in that: The historical settlement unit includes a risk model subunit and a risk assessment subunit; The risk model subunit is used to obtain historical settlement data, perform data preprocessing and feature extraction on the historical settlement data, classify the historical settlement data according to data features, and define primary classification labels, perform in-depth feature mining on different categories of historical settlement data, define secondary classification labels and third-class classification labels respectively, and define transaction risk levels of different transaction customers according to the superposition of classification labels; The risk model subunit is used to divide the historical settlement data and transaction risk levels into a training set, a test set and a validation set, train the machine learning model using the training set, and validate and optimize the machine learning model using the test set and the validation set; The risk assessment subunit is used to obtain the historical settlement data of trading customers, and input the historical settlement data into the machine learning model, obtain and update the risk assessment level of each trading customer, and set corresponding fuel trading upper limits and contractable price ranges for trading customers with different risk assessment levels.

9. A fully digital fuel cost analysis and control platform as claimed in claim 4, characterized in that: The fuel control unit is used to dynamically adjust the fuel strategy according to the real-time fuel information, including: Acquiring real-time fuel information, the real-time fuel information including fuel purchase information, fuel transportation information, fuel storage information and fuel demand information; Obtaining a fuel price according to the fuel purchase information, obtaining a fuel inventory according to the fuel storage information, and obtaining a fuel demand according to the fuel demand information; If the fuel price, fuel inventory or fuel demand exceeds the threshold range, it is sent to the fuel strategy unit, the corresponding value of the benefit function in the fuel strategy unit is updated, and the fuel strategy is re-acquired.

10. A fully digital fuel cost analysis and control platform as claimed in claim 4, characterized in that: The fuel strategy formulated according to the next period fuel price and the next period fuel supply and demand information includes: Obtain historical fuel procurement information, historical fuel transportation information, historical fuel storage information, historical fuel demand information and historical fuel benefit information, obtain a benefit function and a benefit evaluation function based on the historical fuel procurement information, historical fuel transportation information, historical fuel storage information, historical fuel demand information and historical fuel benefit information, input the next period's fuel price and the next period's fuel supply and demand information into the benefit function, continuously and randomly adjust the various fuel quantities in the benefit function, and use the benefit evaluation function to evaluate the fuel benefits generated by each randomly generated fuel strategy until a fuel strategy with the best evaluation result is obtained, and this fuel strategy is used as the final fuel strategy.

Images