Power cost accounting analysis method and system based on multiple factors

Abstract

The invention discloses an electric power cost accounting analysis method and system based on multiple factors, relates to the technical field of electric power economic analysis, and solves the problems that in the prior art, extra costs such as electric power transmission and carbon emission are not considered, and different costs are not effectively predicted. And the accuracy and the efficiency of the electric power cost accounting system are low. Generating a prediction adjustment window after generating a sub-cost according to power plant data and policy data and calculating a total power cost based on the sub-cost; generating a predicted sub-cost according to the prediction adjustment window and the historical sub-cost; an optimization suggestion is generated according to the predicted sub-cost, the power cost is split, a plurality of sub-costs are finely calculated according to factors required by each sub-cost, the power cost in a period of time in the future is predicted, and the model is corrected according to a prediction result, so that the later prediction of the model is more accurate, and the prediction efficiency is improved. The support is provided for the advanced optimization of the electric power cost, and the accuracy and efficiency of the electric power cost accounting system are improved.

Description

Technical Field

[0001] The present application belongs to the technical field of power economic analysis, and specifically is a method and system for power cost accounting and analysis based on multiple factors. Background Art

[0002] Power cost accounting is a complex and meticulous process, involving the collection, classification, aggregation, and accounting of all costs incurred by power companies during production, operations, and management. Power cost accounting helps power companies effectively control expenditures during production and operations. Accurate cost data allows companies to promptly identify and correct cost overruns, thereby reducing capital waste during project construction and ensuring efficient use of funds.

[0003] The prior art (invention patent application with publication number CN117010946A) discloses a production and operation cost accounting system for a thermal power plant and a method for using the same, wherein a production and operation cost accounting system for a thermal power plant includes a central processing module, and the central processing module is signal-connected to a daily fusion transaction auxiliary decision-making module within the region, a comprehensive analysis module for power plant production and operation, a numerical weather forecast module, and a regional supply and demand forecast module; the daily fusion transaction auxiliary decision-making module within the region establishes a fusion transaction market clearing and settlement model based on medium- and long-term daily fusion transaction rules for electricity, collects market transaction data, generates an electricity price change curve, calculates a weighted average transaction price, provides users with a real-time monitoring function, and tracks the daily fusion market clearing volume and market clearing price in real time.

[0004] The above case constructed multiple models for the power plant and calculated its corresponding electricity costs, but failed to consider additional costs such as electricity transportation and carbon emissions, and lacked effective forecasting of different costs. If future costs seriously exceeded the budget, it might cause damage to the company's interests, resulting in low accuracy and efficiency of the electricity cost accounting system. Therefore, the electricity cost accounting system still needs further improvement. Summary of the Invention

[0005] The present application aims to solve at least one of the technical problems existing in the prior art; to this end, the present application proposes a multi-factor based electricity cost accounting analysis method and system to solve the technical problems that the prior art lacks consideration of additional costs such as electricity transportation and carbon emissions, and lacks effective prediction of different costs, resulting in low accuracy and efficiency of the electricity cost accounting system.

[0006] To achieve the above-mentioned objectives, the first aspect of the present application provides a multi-factor-based electricity cost accounting and analysis system, comprising: a data acquisition module, a data analysis module, an early warning module, and a database; the data acquisition module is electrically and / or communicatively connected to the data analysis module; the data analysis module is electrically and / or communicatively connected to the early warning module; the database is electrically and / or communicatively connected to the data acquisition module, the data analysis module, and the early warning module, respectively;

[0007] The data acquisition module acquires power plant data, market data, policy data, and environmental data through data acquisition equipment; the power plant data includes power plant ID, power generation equipment data, and power transmission equipment data; the market data includes unit carbon price; and the environmental data includes weather forecast data;

[0008] The data analysis module generates a plurality of sub-costs based on power plant data, market data, and policy data; calculates the total electricity cost corresponding to the power plant ID based on the plurality of sub-costs and generates a forecast adjustment window; generates a plurality of forecast sub-costs based on the forecast adjustment window, policy data, and a plurality of historical sub-costs; and generates optimization suggestions based on the plurality of forecast sub-costs.

[0009] The early warning module: makes prompts according to the alarm signal and contacts the management personnel;

[0010] The database is used to store data of each module and store historical data required for training the model.

[0011] Through the above steps, this application has carried out a detailed breakdown of electricity costs and accurately measured the key factors of each sub-cost. At the same time, it has further made a forward-looking forecast of future electricity costs. According to the feedback of the forecast results, the model has been timely adjusted and optimized, thereby significantly improving the accuracy of the model's subsequent forecasts, laying a solid foundation for the pre-optimization of electricity costs, and effectively enhancing the accuracy and operating efficiency of the electricity cost accounting system.

[0012] Furthermore, the generation of several sub-costs based on power plant data, market data, and policy data includes:

[0013] Obtaining power plant data, market data, policy data, and environmental data; the power plant data includes power generation equipment data and transmission equipment data; the power generation equipment data includes initial investment, equipment life, unit fuel consumption, power generation, and carbon emissions; the transmission equipment data includes transmission capacity, unit loss cost, and transmission equipment parameters; the market data includes unit carbon price;

[0014] By formula Calculate power cost DC t ; where t represents the number of the time period, CT iIt is expressed as the initial investment amount of the i-th type power supply; SS i It is expressed as the life of the power supply equipment of category i, NLS is expressed as the number of hours of use per year, CB t,i Expressed as the output ratio of the i-th type power supply in the t-th time period; DDRX t,i Expressed as the fuel consumption per unit power generation of the i-th type power source in the t-th time period, RJ t Expressed as fuel price, WC t,i Expressed as the operation and maintenance cost of the i-th type power supply in the t-th time period, FDL t,i It is expressed as the power generation of the i-th type power source in the t-th time period;

[0015] By formula DXC t =α t ×SDL t ×DSC+LHC t Calculating Power Grid System Costs DXC t ; Among them, α t Expressed as loss coefficient, α t ∈(0,1), the loss coefficient is calculated according to the transmission equipment parameters; SDL t It is expressed as the amount of electricity transmitted in the tth time period; DSC is expressed as the unit loss cost; LHC t Expressed as the flexibility cost in the tth time period;

[0016] By formula TPC t =TPL t ×TJ t +ΔZCC t Calculating the cost of carbon emissions (TPC) t Among them, TPL t Expressed as carbon emissions in the tth time period; TJ t Expressed as the unit carbon price in the tth time period; ΔZCC t Expressed as the change in policy cost during the tth time period.

[0017] Furthermore, the loss coefficient is calculated based on the parameters of the power transmission equipment, including:

[0018] Acquiring power transmission equipment parameters and environmental data; the power transmission equipment parameters include resistance and current;

[0019] By formula Calculate the loss coefficient α t ; Among them, α0 represents the reference loss coefficient, α0∈(0,1); R t Expressed as a real-time resistance function, the formula Calculation, R0 represents the wire resistance at standard temperature BW; CW tIt is expressed as the conductor transmission temperature in the tth time period, DW is expressed as unit temperature, k is expressed as the resistance temperature coefficient; I t Expressed as the current corresponding to the tth time period, I e It is represented by the line rated current, γ is represented by the line aging sensitivity coefficient, γ∈(0,1); ST t Indicates the running time of the line, ST design Expressed as the circuit design life.

[0020] Furthermore, the calculation of the total electricity cost corresponding to the power plant ID based on the plurality of sub-costs and the generation of the forecast adjustment window include:

[0021] Obtaining a plurality of sub-costs; the plurality of sub-costs including power supply cost, grid system cost, and carbon emission cost;

[0022] By formula DZB t =DC t +DXC t +TPC t Calculate the total cost of electricity DZB t ;

[0023] Obtain several historical predicted sub-costs corresponding to several sub-costs in the current time period;

[0024] By formula Calculate the prediction error YW corresponding to the sub-cost t,j Among them, YC t,j It is expressed as the predicted cost corresponding to the jth sub-cost in the tth time period, SC t,j It represents the actual cost corresponding to the jth sub-cost in the tth time period; k represents the time window size required to predict the sub-cost;

[0025] A prediction adjustment window is generated according to the prediction error and its corresponding error range; the error range is obtained through the prediction adjustment window corresponding to the sub-cost.

[0026] Furthermore, the error range is obtained by the prediction adjustment window corresponding to the sub-cost, including:

[0027] Obtaining a number of historical forecast errors within a forecast adjustment window corresponding to a number of sub-costs;

[0028] By formula Calculate the historical error mean LWJ t,j ;

[0029] By formula Calculate the historical error standard deviation LWB t,j ;

[0030] Let the error range WF corresponding to the sub-cost bet，j ∈[LWJ t,j -a×LWB t,j , LWJ t,j +a×LWB t,j ]; where a is the confidence level coefficient, a∈(0,1).

[0031] This application calculates several sub-costs in the current time period and their corresponding historical predicted sub-costs to obtain their corresponding multiple prediction errors, and obtains the corresponding several historical error means in the current prediction adjustment window. Since the error range is dynamically adjusted according to the predicted sub-cost and the actual sub-cost, the error range of different sub-costs in different time periods is also adaptively adjusted, which can improve the accuracy of the prediction.

[0032] Furthermore, generating a prediction adjustment window according to the prediction error and its corresponding error range includes:

[0033] Get the prediction error YW t,j Its corresponding error range and the corresponding forecast adjustment window YS in the previous time period t-1,j ; The error range includes the upper limit S of the error range t，j ;

[0034] Calculate the forecast adjustment window YS by formula t，j ;

[0035]

[0036] Among them, TC j,max and TC j,min Represents the maximum and minimum values ​​adjusted for the j-th time window; and They represent the rounding up symbol and the rounding down symbol respectively, max() and min() represent the maximum value and minimum value operations respectively; AF t,j Expressed as safe and stable range, AF t,j ∈[LWJ t,j -b×LWB t,j , LWJ t,j +b×LWB t,j ]; where b is the stability coefficient, b∈(0,1), b <a;ΔC t,j It is expressed as the difference between the maximum value of the adjustment window and the forecast adjustment window corresponding to the previous time period.

[0037] Furthermore, generating a plurality of forecast sub-costs based on the forecast adjustment window, policy data, and a plurality of historical sub-costs includes:

[0038] Obtaining a forecast adjustment window, policy data, power plant data, market data, and environmental data, as well as a number of sub-costs and sub-cost tags; the sub-cost tags include a power source tag, a system tag, and a carbon tag;

[0039] Select several historical policy data, power plant data, market data and environmental data and their corresponding historical sub-costs in the forecast adjustment window corresponding to the sub-cost tag;

[0040] Integrate a number of historical policy data, power plant data, market data and environmental data and their corresponding historical sub-costs into a number of forecast series;

[0041] Input the cost label and several prediction sequences into the cost prediction model to obtain the cost label and its corresponding prediction sub-cost;

[0042] The cost prediction model is constructed through a machine learning model, including:

[0043] Obtain several sub-cost tags and their corresponding historical prediction sequences and historical sub-costs;

[0044] Divide several sub-cost labels and their corresponding historical prediction sequences and historical sub-costs into training data, validation data, and test data; perform data preprocessing on the training data, validation data, and test data to obtain training sets, validation sets, and test sets;

[0045] Select a machine learning model as the base model;

[0046] Train the basic model using the training set, and adjust the learning rate and hyperparameters on the validation set to obtain the pre-trained model;

[0047] By validating the pre-trained model on the test set, we finally obtain a cost prediction model that inputs cost labels and several prediction sequences, and outputs cost labels and their corresponding predicted sub-costs.

[0048] Furthermore, generating optimization suggestions based on the plurality of predicted sub-costs includes:

[0049] Obtain several forecast sub-cost and power plant data;

[0050] The total predicted cost is obtained by summing up several predicted sub-costs;

[0051] Determine whether the predicted total cost is greater than the total cost threshold;

[0052] Yes, generate cost overcomputing warning signals, build a cost optimization function based on power plant data, solve the cost optimization function using a deep reinforcement learning model to obtain the optimal solution, and generate optimization suggestions based on this solution;

[0053] No, when several predicted sub-costs are greater than their corresponding sub-cost thresholds, a sub-cost overcalculation warning signal is generated; several sub-cost optimization functions are constructed based on the power plant data, and the optimal solutions are obtained for these sub-cost optimization functions through a deep reinforcement learning model, and optimization suggestions are generated based on this; otherwise, no action is taken.

[0054] Furthermore, the constructing of a cost optimization function based on power plant data and the constructing of several sub-cost optimization functions based on power plant data include:

[0055] Obtaining power plant data and market data; the power plant data includes power generation equipment data and transmission equipment data; the power generation equipment data includes unit fuel consumption, power generation, and carbon emissions; the transmission equipment data includes power transmission and transmission equipment parameters; the market data includes unit carbon price;

[0056] By formula minC power =∑ j [(DDRX i ×RJ+WC i )×FDL i ]Construct power cost optimization function C power Among them, DDRX i It is represented by the fuel consumption per unit power generation of the i-th type power source, RJ represents the fuel price, WC i Expressed as the operation and maintenance cost of the i-th power supply, FDL i Expressed as the power generation of the i-th type power source;

[0057] By formula minC grid =∑ k (L k ×LD+GC k ×max(k)) to construct the power grid system cost optimization function C grid ; where k is the line number that needs to be modified, L k It refers to the line loss rate, LD refers to the line loss penalty unit price; GC k It refers to the cost required for line transformation; max(k) represents the number of lines that need to be transformed;

[0058] By formula minC carbon =TPL×TJ to construct carbon cost optimization function C carbon ; TPL represents carbon emissions, TJ represents unit carbon price;

[0059] By the formula minC=C power +C grid +C carbon Construct cost optimization function C.

[0060] Another aspect of the present invention provides a method for calculating and analyzing electricity costs based on multiple factors, comprising:

[0061] S0: Obtain power plant data, market data, policy data and environmental data;

[0062] S1: Generate several sub-costs based on power plant data, market data and policy data;

[0063] S2: Calculate the total electricity cost corresponding to the power plant ID based on several sub-costs and generate a forecast adjustment window;

[0064] S3: Generate several forecast sub-costs based on the forecast adjustment window, policy data and several historical sub-costs;

[0065] S4: Generate optimization suggestions based on several predicted sub-costs;

[0066] S5: Prompt according to the alarm signal and contact the management.

[0067] Compared with the prior art, the present invention has the following advantages:

[0068] 1. This application generates several sub-costs based on power plant data, market data and policy data; generates a forecast adjustment window after calculating the total electricity cost corresponding to the power plant ID based on the several sub-costs; generates several predicted sub-costs based on the forecast adjustment window, policy data and several historical sub-costs; generates optimization suggestions based on the several predicted sub-costs, splits the electricity cost, and calculates several sub-costs in detail with the factors required for each sub-cost, and predicts the electricity cost for a period of time in the future. The model is corrected according to the prediction results, so that the model's subsequent predictions are more accurate, which provides support for the early optimization of electricity costs and improves the accuracy and efficiency of the electricity cost accounting system.

[0069] 2. This application obtains the transmission equipment parameters and environmental data required to calculate the grid system cost, adaptively determines the value of the loss coefficient that appears in the grid system cost calculation formula, and considers various factors that affect the transmission efficiency of the transmission line when transmitting power, so that the grid system cost can obtain accurate values ​​according to different time periods and different transmission states, thereby improving the accuracy of the grid system cost and thus improving the accuracy of the electricity cost.

[0070] 3. This application adaptively adjusts the prediction adjustment window size required by the prediction model for the next prediction based on the prediction error and its corresponding error range. According to the different states of the prediction error and the different adjustable ranges of the prediction adjustment window of the prediction sub-cost, the prediction results of each sub-cost can be made more accurate, providing good data support for subsequent optimization suggestions.

[0071] 4. This application uses a cost prediction model to predict each sub-cost in the future time period and compares it with several cost thresholds pre-set by the power plant. Once the cost exceeds the budget, an early warning signal is immediately issued, and the corresponding optimization function is constructed after the early warning. By solving the optimization function, optimization suggestions are obtained, which improves the comprehensiveness of the power cost accounting and analysis system and makes the system more efficient. BRIEF DESCRIPTION OF THE DRAWINGS

[0072] In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, the following briefly introduces the drawings required for use in the embodiments or the description of the prior art. Obviously, the drawings described below are only some embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without any creative work.

[0073] Figure 1 This is a schematic diagram of the principle of a multi-factor based electricity cost accounting and analysis system for this application;

[0074] Figure 2 Generate a flow chart for the optimization suggestions of this application;

[0075] Figure 3 This is a flow chart of a multi-factor based electricity cost accounting and analysis method for this application. DETAILED DESCRIPTION

[0076] The following will clearly and completely describe the technical solutions of this application in conjunction with the embodiments. Obviously, the embodiments described are only a part of the embodiments of this application, not all of them. Based on the embodiments in this application, all other embodiments obtained by ordinary technicians in this field without making creative efforts are within the scope of protection of this application.

[0077] See also Figure 1 In a first aspect, an embodiment of the present application provides a multi-factor-based electricity cost accounting and analysis system, comprising: a data acquisition module, a data analysis module, an early warning module, and a database; the data acquisition module is electrically and / or communicatively connected to the data analysis module; the data analysis module is electrically and / or communicatively connected to the early warning module; the database is electrically and / or communicatively connected to the data acquisition module, the data analysis module, and the early warning module, respectively;

[0078] Data acquisition module: This module acquires power plant data, market data, policy data, and environmental data through data acquisition equipment. Power plant data includes power plant ID, power generation equipment data, and transmission equipment data. Market data includes unit carbon price. Environmental data includes weather forecast data. Data acquisition equipment includes several sensors. Policy data refers to policies issued on power generation and carbon emissions, including electricity price subsidies and carbon emission subsidies.

[0079] Data Analysis Module: Generates several sub-costs based on power plant data, market data, and policy data. Sub-costs refer to the costs that make up the total cost of a power plant, including power supply costs, grid system costs, and carbon emission costs. After calculating the total power cost corresponding to the power plant ID based on the sub-costs, a forecast adjustment window is generated. The forecast adjustment window refers to the window size corresponding to the forecast sequence when predicting the sub-costs. Based on the forecast adjustment window, policy data, and several historical sub-costs, several predicted sub-costs are generated. Predicted sub-costs refer to the estimated value of sub-costs for a period of time in the future. Optimization suggestions are generated based on the predicted sub-costs.

[0080] Early warning module: issues prompts based on alarm signals and contacts management personnel; alarm signals include cost overcalculation warning signals and sub-cost overcalculation warning signals;

[0081] The database is used to store data for each module and store historical data required for training the model.

[0082] In this embodiment, several sub-costs are generated based on power plant data, market data, and policy data, including:

[0083] Obtain power plant data, market data, policy data, and environmental data; power plant data includes power generation equipment data and transmission equipment data; power generation equipment data includes initial investment, equipment life, unit fuel consumption, power generation, and carbon emissions; transmission equipment data includes transmission capacity, unit loss cost, and transmission equipment parameters; market data includes unit carbon price, etc.; policy data includes policy cost changes, etc.

[0084] By formula Calculate power cost DC t ; where t represents the number of the time period, CT i It is expressed as the initial investment amount of the i-th type power supply; SS i It is expressed as the life of the power supply equipment of category i, NLS is expressed as the number of hours of use per year, CB t,i Expressed as the output ratio of the i-th type of power supply in the t-th time period, that is, the ratio of the power generation of different power supply devices to the total power generation; DDRX t,i Expressed as the fuel consumption per unit power generation of the i-th type power source in the t-th time period, RJ t Expressed as fuel price, WCt,i Expressed as the operation and maintenance cost of the i-th type power supply in the t-th time period, FDL t,i It is expressed as the power generation of the i-th type power source in the t-th time period; the initial investment amount of each power source equipment is a fixed value. The more fuel the power generation equipment consumes to produce a certain amount of electricity, and the fuel price gradually increases, the power cost will increase accordingly;

[0085] By formula DXC t =α t ×SDL t ×DSC+LHC t Calculating Power Grid System Costs DXC t ; Among them, α t Expressed as loss coefficient, α t ∈(0,1), the loss coefficient is calculated based on the transmission capacity and transmission equipment parameters; SDL t It is expressed as the amount of electricity transmitted in the tth time period; DSC is expressed as the unit loss cost, and the specific value is set according to experience; LHC t Expressed as the flexibility cost in time period t, flexibility cost refers to the costs incurred in various situations, such as line repair and system maintenance, in addition to the costs consumed by transmission lines. The more power is transmitted, the more severe the line losses are, the higher the unit consumption cost is, and the required grid system cost will also increase accordingly.

[0086] By formula TPC t =TPL t ×TJ t +ΔZCC t Calculating the cost of carbon emissions (TPC) t Among them, TPL t Expressed as carbon emissions in the tth time period; TJ t Expressed as the unit carbon price in the tth time period; ΔZCC t It is expressed as the change in policy cost during the tth time period; when the policy changes, the subsidy amount for power plants' carbon emissions is reduced, and the carbon emissions gradually increase, the carbon emission cost will also increase on the basis of unchanged unit carbon price.

[0087] This embodiment divides various aspects of electricity production costs and calculates the costs in different directions separately, so that the sub-costs in each direction within the time period are more accurate, the accuracy of cost accounting is improved, and good data support is provided for subsequent model corrections and optimization suggestions.

[0088] The loss coefficient in this embodiment is calculated based on the parameters of the power transmission equipment, including:

[0089] Obtain transmission equipment parameters and environmental data; transmission equipment parameters include resistance and current; environmental data include temperature, etc.;

[0090] By formula Calculate the loss coefficient α t ; Among them, α0 represents the reference loss coefficient, α0∈(0,1), the specific value is set according to experience, the reference loss coefficient refers to the reference loss coefficient when the line is in the new line state, and the corresponding loss coefficient will increase with use; R t Expressed as a real-time resistance function, the formula Calculation is performed, where R0 represents the resistance of the wire at the standard temperature BW. In this embodiment, the standard temperature BW is set to 20°C; CW t It represents the wire transmission temperature in the tth time period, DW represents the unit temperature, and the specific value is set according to experience. In this embodiment, DW is set to 1°C, k represents the resistance temperature coefficient, and the specific value is determined by the material properties. For example, the resistance temperature coefficient of copper is about 0.4% / °C; I t Expressed as the current corresponding to the tth time period, I e It is expressed as the line rated current, γ is the line aging sensitivity coefficient, γ∈(0,1), and the specific value is determined according to the material properties, such as 0.015 for copper wire and 0.02 for aluminum wire; ST t Indicates the running time of the line, ST design It is expressed as the design life of the line; as the transmission temperature of the wire increases, the corresponding resistance value of the wire will gradually increase. At the same time, as the current in the line and the length of time the line is used increase, the loss coefficient of the line will also gradually increase.

[0091] This embodiment obtains the transmission equipment parameters and environmental data required to calculate the grid system cost, intelligently and adaptively adjusts the loss coefficient in the grid system cost calculation formula, and analyzes the various factors affecting transmission efficiency faced by transmission lines in different time periods and transmission states during the power transmission process, thereby ensuring that the grid system cost can accurately reflect various actual conditions. This not only improves the accuracy of grid system cost calculation, but also promotes a leap in the overall accuracy of electricity costs, providing solid support for cost optimization and management in the power industry.

[0092] In this embodiment, the total electricity cost corresponding to the power plant ID is calculated based on the multiple sub-costs to generate a forecast adjustment window, including:

[0093] Obtaining several sub-costs; the several sub-costs include power supply cost, grid system cost and carbon emission cost;

[0094] By formula DZB t =DC t +DXCt +TPC t Calculate the total cost of electricity DZB t ;The total cost of electricity increases with the increase of several sub-costs;

[0095] Obtain several historical predicted sub-costs corresponding to several sub-costs in the current time period;

[0096] By formula Calculate the prediction error YW corresponding to the sub-cost t,j Among them, YC t,j It is expressed as the predicted cost corresponding to the jth sub-cost in the tth time period, SC t,j It represents the actual cost corresponding to the jth sub-cost in the tth time period; k represents the time window size required to predict the sub-cost; the larger the time window, the longer the corresponding prediction sequence is, and the more historical data is required; when the predicted sub-cost differs greatly from its corresponding actual sub-cost, it means that the result obtained by the prediction model is not accurate enough, which will result in a large error, and the prediction error will increase accordingly;

[0097] A forecast adjustment window is generated based on the forecast error and its corresponding error range; the error range is obtained through the forecast adjustment window corresponding to the sub-cost.

[0098] The error range in this embodiment is obtained by the prediction adjustment window corresponding to the sub-cost, including:

[0099] Obtaining a number of historical forecast errors within a forecast adjustment window corresponding to a number of sub-costs;

[0100] By formula Calculate the historical error mean LWJ t,j The mean historical error is the average of several historical errors within the current forecast time window. When the historical error is larger, the mean historical error increases while the forecast adjustment window remains unchanged.

[0101] By formula Calculate the historical error standard deviation LWB t,j The larger the standard deviation, the greater the volatility of the prediction error; the smaller the standard deviation, the more stable the model prediction;

[0102] The error range WF corresponding to the sub-cost t，j ∈[LWJ t,j -a×LWB t,j , LWJ t,j +a×LWB t,j ]; where a is the confidence level coefficient, a∈(0, 1), and the specific value is set based on experience. In this embodiment, a is set to 95%.

[0103] Generating a prediction adjustment window based on the prediction error and its corresponding error range in this embodiment includes:

[0104] Obtain the prediction error YW t,j and its corresponding error range, as well as the corresponding prediction adjustment window YS in the previous time period t-1,j ; The error range includes the upper limit S of the error range t，j ;

[0105] Calculate the prediction adjustment window YS through the formula t，j ;

[0106]

[0107] where TC j,max and TC j,min represent the maximum and minimum values adjusted for the j-th time window, that is, the maximum and minimum values of the corresponding prediction adjustment window. The specific values are set according to experience. The maximum and minimum values of the prediction adjustment windows corresponding to different sub-costs are not the same. In this embodiment, the maximum value of the prediction adjustment window corresponding to the power cost is set to 72 hours, and the minimum value is set to 6 hours. The maximum and minimum values of the prediction adjustment windows of other sub-costs can be set according to experience; ​​​​​​​​​​​​​​​​​​​This embodiment uses an adaptive adjustment mechanism based on the prediction error and its corresponding error range to flexibly adjust the time window size of the prediction model in the next prediction. According to the different states of the prediction error and the time window adjustment range allowed by each prediction sub-cost, it realizes the refined optimization of the prediction results of each sub-cost, ensuring that the prediction of each sub-cost can achieve higher accuracy, providing a solid and reliable data foundation for the subsequent optimization suggestions, thereby effectively improving the overall prediction and optimization efficiency.

[0109] In this embodiment, a number of predicted sub-costs are generated based on the predicted adjustment window, policy data, and a number of historical sub-costs, including:

[0110] Obtain forecast adjustment windows, policy data, power plant data, market data, and environmental data, as well as several sub-costs and sub-cost tags; sub-cost tags include power tags, system tags, and carbon tags;

[0111] In the forecast adjustment window corresponding to the sub-cost label, select several historical policy data, power plant data, market data, and environmental data, and their corresponding historical sub-costs. This step allows you to select several forecast sequences corresponding to different sub-costs. The lengths of each forecast sequence do not have to be the same.

[0112] Integrate a number of historical policy data, power plant data, market data and environmental data and their corresponding historical sub-costs into a number of forecast series;

[0113] Input the cost label and several prediction sequences into the cost prediction model to obtain the cost label and its corresponding prediction sub-cost;

[0114] The cost prediction model is constructed through a machine learning model, including:

[0115] Obtain several sub-cost tags and their corresponding historical prediction sequences and historical sub-costs;

[0116] Several sub-cost labels and their corresponding historical prediction sequences and historical sub-costs are divided into training data, validation data, and test data. The training data, validation data, and test data are preprocessed to obtain the training set, validation set, and test set. The ratio of the training set, test set, and validation set is 7:2:1.

[0117] Select a machine learning model as the base model; machine learning models include LSTM models, etc.

[0118] Train the basic model using the training set, and adjust the learning rate and hyperparameters on the validation set to obtain the pre-trained model;

[0119] By validating the pre-trained model on the test set, we finally obtain a cost prediction model that inputs cost labels and several prediction sequences, and outputs cost labels and their corresponding predicted sub-costs.

[0120] See also Figure 2 In this embodiment, the optimization suggestions are generated based on several predicted sub-costs, including:

[0121] Obtain several forecast sub-cost and power plant data;

[0122] The total predicted cost is obtained by summing up several predicted sub-costs;

[0123] Determine whether the predicted total cost is greater than the total cost threshold; the total cost threshold is set based on experience;

[0124] Yes, it generates a cost overcalculation warning signal and constructs a cost optimization function based on the power plant data. The cost optimization function is constructed when the total cost exceeds the budget, and is used to solve the solution that can reduce the total cost to within the total cost threshold. The cost optimization function is solved using a deep reinforcement learning model to obtain the optimal solution, and optimization suggestions are generated based on this solution. The deep reinforcement learning model is a pre-trained model that can effectively obtain the optimal solution of several optimization functions and generate optimization suggestions based on this.

[0125] If the error is no, determine whether several predicted sub-costs are greater than their corresponding sub-cost thresholds, which are set based on experience. If the error is yes, generate a sub-cost overcalculation warning signal. Construct several sub-cost optimization functions based on the power plant data, solve these sub-cost optimization functions using a deep reinforcement learning model to obtain the optimal solution, and generate optimization recommendations based on this. If the error is no, do nothing.

[0126] In this embodiment, the cost optimization function is constructed based on the power plant data, and several sub-cost optimization functions are constructed based on the power plant data, including:

[0127] Obtain power plant data and market data; power plant data includes power generation equipment data and transmission equipment data; power generation equipment data includes unit fuel consumption, power generation, and carbon emissions; transmission equipment data includes transmission capacity and transmission equipment parameters; market data includes unit carbon price;

[0128] By formula minC power =∑ j [(DDRX i ×RJ+WC i )×FDL i ]Construct power cost optimization function C power Among them, DDRX i It is represented by the fuel consumption per unit power generation of the i-th type power source, RJ represents the fuel price, WC iExpressed as the operation and maintenance cost of the i-th power supply, FDL i Expressed as the power generation of the i-th type power source;

[0129] By formula minC grid =∑ k (L k ×LD+GC k ×max(k)) to construct the power grid system cost optimization function C grid ; where k is the line number that needs to be modified, L k It refers to the line loss rate, LD refers to the line loss penalty unit price; GC k It refers to the cost required for line transformation; max(k) represents the number of lines that need to be transformed;

[0130] By formula minC carbon =TPL×TJ to construct carbon cost optimization function C carbon ; TPL represents carbon emissions, TJ represents unit carbon price;

[0131] By the formula minC=C power +C grid +C carbon Construct a cost optimization function C; in the cost optimization function, there are several optimization conditions when optimizing, such as, C power ≤YC power , C grid ≤YC grid , C carbon ≤YC carbon ; Among them YC power Denoted as the power cost threshold, YC grid Expressed as the grid system cost threshold, YC carbon Expressed as carbon emission cost threshold.

[0132] This embodiment uses a cost prediction model to make detailed estimates of various sub-costs in future time periods and carefully compares these predicted values ​​with cost thresholds. Once any cost item is found to exceed the preset budget range, the system will immediately trigger the early warning mechanism and quickly issue an alarm. After that, the system will specifically construct a corresponding optimization mathematical model, solve the model, and generate a series of practical cost optimization strategies, thereby improving the overall performance of the electricity cost accounting and analysis system and significantly enhancing its operational efficiency.

[0133] See also Figure 3 Another aspect of the present application provides a method for analyzing electricity cost calculation based on multiple factors, including:

[0134] S0: Obtain power plant data, market data, policy data and environmental data;

[0135] S1: Generate several sub-costs based on power plant data, market data and policy data;

[0136] S2: Calculate the total electricity cost corresponding to the power plant ID based on several sub-costs and generate a forecast adjustment window;

[0137] S3: Generate several forecast sub-costs based on the forecast adjustment window, policy data and several historical sub-costs;

[0138] S4: Generate optimization suggestions based on several predicted sub-costs;

[0139] S5: Prompt according to the alarm signal and contact the management.

[0140] Some of the data in the above formula are calculated by removing the dimensions and taking their numerical values. The formula is a formula that is closest to the actual situation obtained by software simulation of a large amount of collected data; the preset parameters and preset thresholds in the formula are set by technical personnel in this field according to actual conditions or obtained through simulation of a large amount of data.

[0141] The working principle of this application is: by obtaining power plant data, market data, policy data and environmental data; generating several sub-costs based on the power plant data, market data and policy data; generating a forecast adjustment window after calculating the total electricity cost corresponding to the power plant ID based on the several sub-costs; generating several predicted sub-costs based on the forecast adjustment window, policy data and several historical sub-costs; generating optimization suggestions based on several predicted sub-costs; making prompts based on alarm signals and contacting management personnel to split the electricity cost, and finely calculate several sub-costs based on the factors required for each sub-cost, and at the same time predict the electricity cost for a period of time in the future, and correcting the model based on the prediction results, so that the model's subsequent predictions are more accurate, providing support for the early optimization of electricity costs, improving the accuracy and efficiency of the electricity cost accounting system, and avoiding the problem that the existing technology lacks consideration of additional costs such as electricity transportation and carbon emissions, and lacks effective prediction of different costs, resulting in low accuracy and efficiency of the electricity cost accounting system.

[0142] The above embodiments are only used to illustrate the technical method of the present application and are not intended to limit it. Although the present application has been described in detail with reference to the preferred embodiments, those skilled in the art should understand that the technical method of the present application can be modified or replaced by equivalents without departing from the spirit and scope of the technical method of the present application.

Claims

1. A multi-factor based electricity cost accounting and analysis system, characterized in that: include: Interconnected data acquisition module and data analysis module; The data acquisition module acquires power plant data, market data, policy data and environmental data through data acquisition equipment; The data analysis module generates several sub-costs based on power plant data, market data, and policy data; generates a forecast adjustment window after calculating the total electricity cost corresponding to the power plant ID based on the several sub-costs; generates several predicted sub-costs based on the predicted adjustment window, policy data, and several historical sub-costs; and generates optimization suggestions based on the several predicted sub-costs.

2. The multi-factor-based electricity cost accounting and analysis system according to claim 1, characterized in that: The method generates several sub-costs based on power plant data, market data, and policy data, including: Obtaining power plant data, market data, policy data, and environmental data; the power plant data includes power generation equipment data and transmission equipment data; the power generation equipment data includes initial investment, equipment life, unit fuel consumption, power generation, and carbon emissions; the transmission equipment data includes transmission capacity, unit loss cost, and transmission equipment parameters; the market data includes unit carbon price; By formula Calculate power cost DC t ; where t represents the number of the time period, CT i It is expressed as the initial investment amount of the i-th type power supply; SS i It is expressed as the life of the power supply equipment of category i, NLS is expressed as the number of hours of use per year, CB t,i Expressed as the output ratio of the i-th type power supply in the t-th time period; DDRX t,i Expressed as the fuel consumption per unit power generation of the i-th type power source in the t-th time period, RJ t Expressed as fuel price, WC t,i Expressed as the operation and maintenance cost of the i-th type power supply in the t-th time period, FDL t,i It is expressed as the power generation of the i-th type power source in the t-th time period; By formula DXC t =α t ×SDL t ×DSC+LHC t Calculating Power Grid System Costs DXC t ; Among them, α t Expressed as loss coefficient, α t ∈(0,1), the loss coefficient is calculated according to the transmission equipment parameters; SDL t It is expressed as the amount of electricity transmitted in the tth time period; DSC is expressed as the unit loss cost; LHC t Expressed as the flexibility cost in the tth time period; By formula TPC t =TPL t ×TJ t +ΔZCC t Calculating the cost of carbon emissions (TPC) t Among them, TPL t Expressed as carbon emissions in the tth time period; TJ t Expressed as the unit carbon price in the tth time period; ΔZCC t Expressed as the change in policy cost during the tth time period.

3. The multi-factor-based electricity cost accounting and analysis system according to claim 2, characterized in that: The loss coefficient is calculated based on the transmission equipment parameters, including: Acquiring power transmission equipment parameters and environmental data; the power transmission equipment parameters include resistance and current; By formula Calculate the loss coefficient α t ; Among them, α0 represents the reference loss coefficient, α0∈(0,1); R t Expressed as a real-time resistance function, the formula Calculation, R0 represents the wire resistance at standard temperature BW; CW t It is expressed as the conductor transmission temperature in the tth time period, DW is expressed as unit temperature, k is expressed as the resistance temperature coefficient; I t Expressed as the current corresponding to the tth time period, I e It is represented by the line rated current, γ is represented by the line aging sensitivity coefficient, γ∈(0,1); ST t Indicates the running time of the line, ST design Expressed as the circuit design life.

4. The power cost accounting and analysis system based on multiple factors according to claim 1, characterized in that: The step of calculating the total electricity cost corresponding to the power plant ID based on the plurality of sub-costs and then generating a forecast adjustment window includes: Obtaining a plurality of sub-costs; the plurality of sub-costs including power supply cost, grid system cost, and carbon emission cost; The total cost of electricity is calculated by summing up several sub-costs; Obtain several historical forecast sub-costs corresponding to several sub-costs in the current time period; By formula Calculate the prediction error YW corresponding to the sub-cost t,j Among them, YC t,j It is expressed as the predicted cost corresponding to the jth sub-cost in the tth time period, SC t,j It represents the actual cost corresponding to the jth sub-cost in the tth time period; k represents the time window size required to predict the sub-cost; A prediction adjustment window is generated according to the prediction error and its corresponding error range; the error range is obtained through the prediction adjustment window corresponding to the sub-cost.

5. The multi-factor-based electricity cost accounting and analysis system according to claim 4, characterized in that: The error range is obtained through the forecast adjustment window corresponding to the sub-cost, including: Obtaining a number of historical forecast errors within a forecast adjustment window corresponding to a number of sub-costs; By formula Calculate the historical error mean LWJ t,j ; By formula Calculate the historical error standard deviation LWB t,j ; Let the error range WF corresponding to the sub-cost be t，j ∈[LWJ t,j -a×LWB t,j , LWJ t,j +a×LWB t,j ]; where a is the confidence level coefficient, a∈(0,1).

6. The multi-factor-based electricity cost accounting and analysis system according to claim 4, characterized in that: Generating a prediction adjustment window according to the prediction error and its corresponding error range includes: Get the prediction error YW t,j Its corresponding error range and the corresponding forecast adjustment window YS in the previous time period t-1,j ; The error range includes the upper limit S of the error range t，j ; Calculate the forecast adjustment window YS by formula t，j ; Among them, TC j,max and TC j,min Represents the maximum and minimum values ​​adjusted for the j-th time window; and They represent the rounding up symbol and the rounding down symbol respectively, max() and min() represent the maximum value and minimum value operations respectively; AF t,j Expressed as safe and stable range, AF t,j ∈[LWJ t,j -b×LWB t,j , LWJ t,j +b×LWB t,j ]; where b is the stability coefficient, b∈(0,1), b <a;ΔC t,j It is expressed as the difference between the maximum value of the adjustment window and the forecast adjustment window corresponding to the previous time period.

7. The multi-factor-based electricity cost accounting and analysis system according to claim 1, characterized in that: The generating of a plurality of forecast sub-costs according to the forecast adjustment window, policy data and a plurality of historical sub-costs includes: Obtaining a forecast adjustment window, policy data, power plant data, market data, and environmental data, as well as a number of sub-costs and sub-cost tags; the sub-cost tags include a power source tag, a system tag, and a carbon tag; Select several historical policy data, power plant data, market data and environmental data and their corresponding historical sub-costs in the forecast adjustment window corresponding to the sub-cost tag; Integrate a number of historical policy data, power plant data, market data and environmental data and their corresponding historical sub-costs into a number of forecast series; Input the cost label and several prediction sequences into the cost prediction model to obtain the cost label and its corresponding prediction sub-cost; The cost prediction model is constructed through a machine learning model, including: Obtain several sub-cost tags and their corresponding historical prediction sequences and historical sub-costs; Divide several sub-cost labels and their corresponding historical prediction sequences and historical sub-costs into training data, validation data, and test data; perform data preprocessing on the training data, validation data, and test data to obtain training sets, validation sets, and test sets; Select a machine learning model as the base model; Train the basic model using the training set, and adjust the learning rate and hyperparameters on the validation set to obtain the pre-trained model; By validating the pre-trained model on the test set, we finally obtain a cost prediction model that inputs cost labels and several prediction sequences, and outputs cost labels and their corresponding predicted sub-costs.

8. The multi-factor-based electricity cost accounting and analysis system according to claim 1, characterized in that: The optimization suggestions are generated based on several predicted sub-costs, including: Obtain several forecast sub-cost and power plant data; The total predicted cost is obtained by summing up several predicted sub-costs; Determine whether the predicted total cost is greater than the total cost threshold; Yes, generate cost overcomputing warning signals, build a cost optimization function based on power plant data, solve the cost optimization function using a deep reinforcement learning model to obtain the optimal solution, and generate optimization suggestions based on this solution; No, when several predicted sub-costs are greater than their corresponding sub-cost thresholds, a sub-cost overcalculation warning signal is generated; several sub-cost optimization functions are constructed based on the power plant data, and the optimal solutions are obtained for these sub-cost optimization functions through a deep reinforcement learning model, and optimization suggestions are generated based on this; otherwise, no action is taken.

9. The multi-factor-based electricity cost accounting and analysis system according to claim 8, characterized in that: The constructing of a cost optimization function based on power plant data and the constructing of a plurality of sub-cost optimization functions based on power plant data include: Obtaining power plant data and market data; the power plant data includes power generation equipment data and transmission equipment data; the power generation equipment data includes unit fuel consumption, power generation, and carbon emissions; the transmission equipment data includes power transmission and transmission equipment parameters; the market data includes unit carbon price; By formula minC power =∑ i [(DDRX i ×RJ+WC i )×FDL i ]Construct power cost optimization function C power Among them, DDRX i It is represented by the fuel consumption per unit power generation of the i-th type power source, RJ represents the fuel price, WC i Expressed as the operation and maintenance cost of the i-th power supply, FDL i Expressed as the power generation of the i-th type power source; By formula minC grid =∑ k (L k ×LD+GC k ×max(k)) to construct the power grid system cost optimization function C grid ; where k is the line number that needs to be modified, L k It refers to the line loss rate, LD refers to the line loss penalty unit price; GC k It refers to the cost required for line transformation; max(k) represents the number of lines that need to be transformed; By formula minC carbon =TPL×TJ to construct carbon cost optimization function C carbon ; TPL represents carbon emissions, TJ represents unit carbon price; By the formula minC=C power +C grid +C carbon Construct a cost optimization function C.

10. A method for calculating and analyzing electricity cost based on multiple factors, applied to a system for calculating and analyzing electricity cost based on multiple factors according to any one of claims 1 to 9, characterized in that: include: S0: Obtain power plant data, market data, policy data and environmental data; S1: Generate several sub-costs based on power plant data, market data and policy data; S2: Calculate the total electricity cost corresponding to the power plant ID based on several sub-costs and generate a forecast adjustment window; S3: Generate several forecast sub-costs based on the forecast adjustment window, policy data and several historical sub-costs; S4: Generate optimization suggestions based on several predicted sub-costs.

Images