Pumped storage power station risk assessment system and method based on multi-factor interaction

Abstract

The invention provides a pumped storage power station risk assessment system and method based on multi-factor interaction, and the method comprises the steps: obtaining first related data through an input terminal, and transmitting the first related data to a processing terminal; the processing terminal receives the first related data and obtains a preset reference value corresponding to each risk point based on the first related data; the first related data comprises a plurality of risk points; constructing a first model about the pumped storage power station based on a path analysis method or a BP neural network or system dynamics, and obtaining a risk assessment result; sending the risk assessment result to a display device; the display equipment is configured to generate a target picture according to the risk assessment result, and the target picture is displayed in a display of the display equipment. According to the method, accurate prediction of various risks can be realized, the risks of the transaction center can be divided into low, medium and high levels through the calculated comprehensive risk score, and a scientific basis is provided for risk early warning and management decision making of the transaction center.

Description

Technical Field

[0001] The present application relates to the field of electric power technology, and in particular to a risk assessment system and method for a pumped-storage power station based on multi-factor interaction. Background Art

[0002] With the increasing global attention to environmental issues and climate change, renewable energy such as hydropower plays an increasingly important role in energy supply. Pumped-storage power stations are an effective way to store and dispatch energy, which can solve the instability caused by the volatility and intermittency of renewable energy and improve the sustainability of renewable energy. There is a balance problem between energy supply and demand in the power system, especially with the rapid growth of electricity demand. As a flexible power peak-shaving facility, pumped-storage power stations can reserve energy when there is an excess of energy supply and release hydropower for power generation during peak periods, thereby balancing the supply and demand of electricity. The stability of the power system is crucial to ensure the stable operation of energy supply and social economy. Pumped-storage power stations can serve as a backup power source for the power system, provide stable power output, and enhance the reliability of the power system. Countries around the world are actively promoting sustainable development, including reducing the use of fossil fuels and increasing the proportion of renewable energy. Pumped-storage power stations, as a clean energy storage technology, are in line with the concept of sustainable development. Many countries are undergoing energy transformation, promoting power market reform, and increasing the proportion of renewable energy. Pumped-storage power stations can promote the large-scale application of renewable energy and help the smooth progress of energy transformation and market reform.

[0003] Risk assessment is also required when building and operating a pumped storage power station. Existing risk management methods mostly analyze and predict a single type of risk, and these methods are limited in their effectiveness when dealing with the interaction between different risks. At the same time, in the process of data collection and use, problems such as missing data and anomalies are often encountered, which affect the reliability of the analysis model and the accuracy of the results. Summary of the invention

[0004] The present application provides a pumped-storage power station risk assessment system and method based on multi-factor interaction to achieve accurate risk prediction and correlation analysis and improve user experience.

[0005] In a first aspect, the present application provides a pumped storage power station risk assessment system based on multi-factor interaction, comprising: an input terminal, a processing terminal and a display device; the input terminal and the display device are respectively connected to a communication interface of the processing terminal;

[0006] The input terminal is configured to: obtain first relevant data and send it to the processing terminal;

[0007] The processing terminal is configured to:

[0008] receiving the first relevant data, and acquiring a preset reference value corresponding to each risk point based on the first relevant data; the first relevant data includes a plurality of risk points;

[0009] Constructing a first model of the pumped storage power station based on a path analysis method or a BP neural network or system dynamics, and obtaining a risk assessment result;

[0010] Sending the risk assessment result to the display device;

[0011] The display device is configured to generate a target image according to the risk assessment result, and display the target image on a display of the display device.

[0012] In some possible implementations, the processing terminal includes a storage unit, a computing unit, and a control unit, wherein the storage unit and the computing unit are electrically connected to the control unit respectively; the storage unit is used to store the calculated intermediate values ​​of the evaluation system; the computing unit is used to perform solution calculations; and the control unit is used to send control instructions.

[0013] In some possible implementations, the first relevant data includes historical data on investment and cost risk points, such as historical new investments and scrapped assets, and historical operation and maintenance costs.

[0014] In some possible implementations, when a risk assessment result is obtained based on a path analysis method, after receiving the first relevant data, the method includes:

[0015] Generate a multi-factor variation function based on the first relevant data fitting as a risk key point benchmark value;

[0016] Monitor the actual value of each risk point and calculate the difference between the actual value and the benchmark value;

[0017] A path analysis method is introduced to analyze and verify the degree of influence of factors based on the difference value, to construct a first model, and to obtain a risk assessment result based on the first model.

[0018] In some possible implementations, generating a multi-factor change function based on the first related data includes:

[0019] When the first correlation data is linear data, selecting a univariate or multivariate linear fitting function to generate;

[0020] When the first correlation data is nonlinear data, polynomial fitting, exponential fitting or other nonlinear fitting functions are selected.

[0021] In some possible implementations, the risk assessment results obtained based on the BP neural network include:

[0022] Normalize historical cost data, historical new investment and other risk points and their related variable data;

[0023] Dividing the risk point variable data and the corresponding influencing variable data into multiple data subsets, wherein one subset of the multiple data subsets is used as a validation set, and the remaining subsets are used as training sets;

[0024] Before training the neural network, a genetic algorithm is used to optimize the network structure and initial weights;

[0025] Constructing the BP neural network, the BP neural network includes an input layer, a hidden layer and an output layer, wherein the number of the input layer is determined according to the influencing path and the risk point and its motivation, the number and structure of the hidden layer are determined according to the genetic algorithm, and the output of the output layer is the investment return risk prediction value;

[0026] Using the training set, a back propagation algorithm is applied to adjust the network weights and thresholds of the BP neural network to obtain the first model.

[0027] In some possible implementations, obtaining the investment return risk probability prediction model further includes:

[0028] Calculating the prediction error of the trained BP neural network using the verification set;

[0029] The real-time impact indicator is input into the first model, and a risk assessment result is obtained based on the first model.

[0030] In some possible implementations, when the risk assessment result is obtained based on system dynamics, after receiving the first relevant data, the method includes:

[0031] Input the initial investment of the power plant and historical operation and maintenance cost parameters, and build the power plant investment, cost and pricing model based on the risk points and their related variable fitting results to calculate the capacity price benchmark value of the pumped storage power plant;

[0032] Based on the probability prediction results of the power plant risk points and related variables, combined with the driving relationship between the risk points and related variables, the first model is constructed;

[0033] The capacity electricity price and its probability distribution output by the first model are compared with the benchmark value to obtain the risk assessment result.

[0034] In some possible implementations, based on the probability prediction results of the risk points of the power station and the levels of related variables, combined with the driving relationship between the risk points and the related variables, the first model is constructed including:

[0035] Based on the investment, cost and pricing model of the pumped storage power station, combined with the probability prediction results of the risk points and their related variables, a probability function model of the corresponding variables is constructed;

[0036] The probability model of the risk point and its related variables is introduced into the corresponding formula in the investment, cost and pricing model of the pumped storage power station, and the first model is constructed through system dynamics.

[0037] In a second aspect, the present application provides a pumped storage power station risk assessment method based on multi-factor interaction, which is configured in the pumped storage power station risk assessment system based on multi-factor interaction described in the first aspect, and the method comprises:

[0038] The input terminal obtains the first relevant data and sends it to the processing terminal;

[0039] The processing terminal receives the first relevant data, and acquires a preset reference value corresponding to each risk point based on the first relevant data; the first relevant data includes a plurality of risk points;

[0040] Constructing a first model of the pumped storage power station based on a path analysis method or a BP neural network or system dynamics, and obtaining a risk assessment result;

[0041] Sending the risk assessment result to a display device;

[0042] The display device generates a target image according to the risk assessment result, and displays the target image on a display of the display device.

[0043] It can be seen from the above technical scheme that the present application provides a pumped storage power station risk assessment system and method based on multi-factor interaction, the method comprising an input terminal acquiring first relevant data and sending it to a processing terminal; the processing terminal receiving the first relevant data, and, based on the first relevant data, acquiring a preset reference value corresponding to each risk point; the first relevant data includes multiple risk points; constructing a first model about the pumped storage power station based on a path analysis method or a BP neural network or system dynamics, and obtaining a risk assessment result; sending the risk assessment result to a display device; the display device is configured to generate a target image according to the risk assessment result, and display the target image on the display of the display device. The present invention can not only realize the accurate prediction of multiple risks, but also divide the risks of the trading center into low, medium and high levels through the calculated comprehensive risk score, providing a scientific basis for the risk warning and management decision of the trading center. BRIEF DESCRIPTION OF THE DRAWINGS

[0044] In order to more clearly illustrate the technical solution of the present application, the drawings required for use in the embodiments are briefly introduced below. Obviously, for ordinary technicians in this field, other drawings can be obtained based on these drawings without any creative work.

[0045] Figure 1 A schematic diagram of the structure of a pumped-storage power station risk assessment system based on multi-factor interaction provided in an embodiment of the present application;

[0046] Figure 2 A schematic diagram of constructing a simulation model for system dynamics of an embodiment of the present application;

[0047] Figure 3 This is a flow chart of a risk assessment method for a pumped-storage power station based on multi-factor interaction according to an embodiment of the present application. DETAILED DESCRIPTION

[0048] The following embodiments are described in detail, and examples thereof are shown in the accompanying drawings. When the following description refers to the drawings, unless otherwise indicated, the same numbers in different drawings represent the same or similar elements. The implementations described in the following embodiments do not represent all implementations consistent with the present application. They are only examples of systems and methods consistent with some aspects of the present application as detailed in the claims.

[0049] With the increasing global attention to environmental issues and climate change, renewable energy such as hydropower plays an increasingly important role in energy supply. Pumped-storage power stations are an effective way to store and dispatch energy, which can solve the instability caused by the volatility and intermittency of renewable energy and improve the sustainability of renewable energy. There is a balance problem between energy supply and demand in the power system, especially with the rapid growth of electricity demand. As a flexible power peak-shaving facility, pumped-storage power stations can reserve energy when there is an excess of energy supply and release hydropower for power generation during peak periods, thereby balancing the supply and demand of electricity. The stability of the power system is crucial to ensure the stable operation of energy supply and social economy. Pumped-storage power stations can serve as a backup power source for the power system, provide stable power output, and enhance the reliability of the power system. Countries around the world are actively promoting sustainable development, including reducing the use of fossil fuels and increasing the proportion of renewable energy. Pumped-storage power stations, as a clean energy storage technology, are in line with the concept of sustainable development. Many countries are undergoing energy transformation, promoting power market reform, and increasing the proportion of renewable energy. Pumped-storage power stations can promote the large-scale application of renewable energy and help the smooth progress of energy transformation and market reform.

[0050] Risk assessment is also required when building and operating a pumped storage power station. Existing risk management methods mostly analyze and predict a single type of risk, and these methods are limited in their effectiveness when dealing with the interaction between different risks. At the same time, in the process of data collection and use, problems such as missing data and anomalies are often encountered, which affect the reliability of the analysis model and the accuracy of the results.

[0051] Based on this, the present application provides a pumped-storage power station risk assessment system and method based on multi-factor interaction, which can not only realize accurate prediction of various risks, but also divide the risks of the trading center into low, medium and high levels through the calculated comprehensive risk score, providing a scientific basis for the risk warning and management decision-making of the trading center.

[0052] In some embodiments, Figure 1 As shown, the present application provides a pumped storage power station risk assessment system based on multi-factor interaction, comprising: an input terminal, a processing terminal and a display device; the input terminal and the display device are respectively connected to the communication interface of the processing terminal;

[0053] The input terminal is configured to: obtain first relevant data and send it to the processing terminal;

[0054] The processing terminal is configured to:

[0055] receiving the first relevant data, and acquiring a preset reference value corresponding to each risk point based on the first relevant data; the first relevant data includes a plurality of risk points;

[0056] Constructing a first model of the pumped storage power station based on a path analysis method or a BP neural network or system dynamics, and obtaining a risk assessment result;

[0057] Sending the risk assessment result to the display device;

[0058] The display device is configured to generate a target image according to the risk assessment result, and display the target image on a display of the display device.

[0059] In some embodiments, the processing terminal includes a storage unit, a computing unit and a control unit, wherein the storage unit and the computing unit are electrically connected to the control unit respectively; the storage unit is used to store the calculated intermediate values ​​of the evaluation system; the computing unit is used to perform solution calculations; and the control unit is used to send control instructions.

[0060] In some embodiments, the first relevant data includes historical data on investment and cost risk points, such as historical new investments and scrapped assets, and historical operation and maintenance costs.

[0061] In some embodiments, when the risk assessment result is obtained based on the path analysis method, after receiving the first relevant data, the method includes:

[0062] Generate a multi-factor variation function based on the first relevant data fitting as a risk key point benchmark value;

[0063] Monitor the actual value of each risk point and calculate the difference between the actual value and the benchmark value;

[0064] A path analysis method is introduced to analyze and verify the degree of influence of factors based on the difference value, to construct a first model, and to obtain a risk assessment result based on the first model.

[0065] In some embodiments, generating a multi-factor variation function based on the first correlation data includes:

[0066] When the first correlation data is linear data, selecting a univariate or multivariate linear fitting function to generate;

[0067] When the first correlation data is nonlinear data, polynomial fitting, exponential fitting or other nonlinear fitting functions are selected.

[0068] Specifically, based on the first relevant data, the key points of investment and return risks of pumped storage power stations are determined, and an investment and return risk factor analysis system is constructed. First, historical data of investment and cost risk points in pumped storage power stations, such as historical new investments and scrapped assets, historical operation and maintenance costs, are collected, and the function of each factor change is fitted based on the collected data as the benchmark value of the key risk points; secondly, the actual implementation of each risk point is monitored, such as new investments, costs of each subject, cost progress, etc., and the difference between the actual value and the benchmark value is calculated; finally, the path analysis method is introduced to analyze and verify the degree of influence of factors, and a pumped storage power station investment and return risk factor analysis system is constructed.

[0069] 1. When collecting historical data on investment return risk points of pumped storage power stations and fitting the benchmark values ​​of production risk points, the fitting function theory can be used. The fitting function is a mathematical method that finds a mathematical function to approximate these data points through the data points in a given data set. Its theoretical basis is based on the principle of least squares. The principle of least squares aims to find a straight line or curve that minimizes the sum of the square errors with the given data points. By minimizing the error, the fitting function can effectively approximate the trends and patterns in the data. There are many types of fitting functions, including linear fitting, polynomial fitting, and exponential fitting. Each type of fitting function has its own unique advantages and disadvantages and is suitable for different data types and modeling purposes.

[0070] First, an appropriate fitting function should be selected based on the data type and its general trend of change. For linear data, a univariate or multivariate linear fitting function is usually selected; for nonlinear data, a polynomial fitting, exponential fitting or other nonlinear fitting functions are usually selected.

[0071] 1) Linear fitting function is a common fitting function, which assumes that the data points are linearly distributed in the Cartesian coordinate system. It can be specifically divided into univariate linear function and multivariate linear function.

[0072] Univariate linear fitting refers to fitting a set of one-dimensional data points to obtain a straight line. The general form of the straight line equation is:

[0073] y=mx+b;

[0074] Here, m represents the slope and b represents the intercept.

[0075] The univariate linear fitting can be obtained through the principle of least squares method. The principle of least squares method and the derivation process of fitting function are shown in the following formula.

[0076] min∑[y i -(mx i +b)] 2 ;

[0077]

[0078] Multivariate linear fitting refers to fitting a set of multidimensional data points to obtain a hyperplane. The general form of the hyperplane equation is:

[0079] y=b+ω1x1+ω2x2+…+ω n x n ;

[0080] Where b represents the intercept, ω i represents the weight of the ith independent variable.

[0081] The multivariate linear fitting function can be obtained by the orthogonalization method. The basic idea is to orthogonalize the independent variable matrix and then use the orthogonal matrix to solve the equation system. The steps are as follows:

[0082] a. Orthogonalize the independent variable matrix X to obtain the orthogonal matrix Q and the diagonal matrix Λ;

[0083] b. Calculate the orthogonalized independent variable matrix Z = Q T X;

[0084] c. Solve the equation system ZW=Q T Y;

[0085] d. Calculate the weight matrix W = Z -1 Q T Y;

[0086] e. Calculate the intercept vector B = Y-XW.

[0087] 2) Nonlinear fitting functions include polynomial fitting functions, exponential fitting functions, etc. Polynomial fitting is a nonlinear fitting method used to fit data with nonlinear relationships. The general form of a polynomial function is:

[0088] f(x)=a0+a1x+a2x 2 +…+a n x n ;

[0089] In the formula, a0, a1, a2, ... a n are the coefficients of the polynomial function.

[0090] Lagrange interpolation is a polynomial fitting method that constructs a polynomial function that passes through all data points given a set of data points. The general form of the Lagrange interpolation polynomial is:

[0091]

[0092] In the formula, y i is the data point x i The corresponding vertical coordinate; l i (x) is the Lagrangian basis function, defined as:

[0093]

[0094] Exponential fitting is used to fit data with an exponential relationship. The general form of the exponential function is:

[0095] f(x)=a·b x ;

[0096] Where a and b are the coefficients of the exponential function.

[0097] Nonlinear least squares is a method used to determine the coefficients of an exponential function. It determines the coefficients by minimizing the sum of squares of a fitting function. The fitting function is defined as:

[0098]

[0099] 2. Monitor the actual implementation of each risk point and calculate the actual difference level based on the risk point benchmark value. Take the investment in new assets and the cost progress of each account as an example to introduce the monitoring method.

[0100] In terms of new asset investment, the progress of new asset investment is reflected through two indicators: new investment deviation value and new investment deviation ratio.

[0101] ΔIy =I y -I y ′;

[0102]

[0103] In the formula, ΔI y represents the deviation value of new investment; β y Indicates the deviation ratio of new investment; I y Indicates the actual value of new asset investment; I y ′ represents the benchmark value of new asset investment.

[0104] In terms of the cost progress of each subject, we first count the actual value of the cost of each subject, the actual value of the production cost of each subject, and the actual value of the management fee of each subject reported.

[0105] Secondly, the progress of different subject costs, production costs and administrative expenses is reflected through the monthly cumulative completion ratio of different subject costs, monthly cumulative completion ratio of production costs and monthly cumulative completion ratio of administrative expenses.

[0106] 3. Based on the calculation of the baseline value of each risk point, the monitoring of the actual value and the difference level between the two, the path analysis method is introduced to explore the direct and indirect impact relationship between the risk point elements and related factors, and to construct a risk point motivation system for investment returns of pumped storage power stations.

[0107] In path analysis, in order to distinguish variables with different characteristics, exogenous variables, endogenous variables and final result variables are used. Exogenous variables refer to those variables in the model that only affect other variables and are not affected by other variables. Endogenous variables are just the opposite of exogenous variables, and refer to variables in the model that can both affect other variables and be affected by other variables. If an endogenous variable is only affected by other variables and has no effect on other variables at all, such an endogenous variable is called a final result variable. Taking a pumped storage power station as an example, the initial investment variable affects other variables such as annual operation and maintenance costs and principal and interest payments, but is not affected by other variables, and is an exogenous variable.

[0108] The basic steps of path analysis are as follows:

[0109] Model setting. The initial theoretical model is set based on the results of previous research and professional background knowledge. The model should include various possible paths, and the various relationships in the model can be clearly described in the form of a path diagram.

[0110] Model identification and model estimation. This step should attempt to estimate the parameters in the model. If the model is set incorrectly, the entire model may not be estimated. At this time, the model should be modified until a preliminary estimate is obtained.

[0111] Model evaluation. Evaluate whether the association of each path is statistically significant and whether it is well consistent with professional knowledge. If it is not statistically significant, you can consider deleting the path; if the fitting result does not conform to professional knowledge, you need to consider whether there are major problems with the entire model framework.

[0112] Model modification. Simplify and improve the model based on the results of model evaluation, and finally obtain a very concise path analysis model that is consistent with professional knowledge and the characteristics of the data. Obviously, a final model can only be obtained after multiple attempts and modifications. After obtaining the path analysis model, the risk assessment results are output.

[0113] In some embodiments, obtaining risk assessment results based on the BP neural network includes:

[0114] Normalize historical cost data, historical new investment and other risk points and their related variable data;

[0115] Dividing the risk point variable data and the corresponding influencing variable data into multiple data subsets, wherein one subset of the multiple data subsets is used as a validation set, and the remaining subsets are used as training sets;

[0116] Before training the neural network, a genetic algorithm is used to optimize the network structure and initial weights.

[0117] Constructing the BP neural network, the BP neural network includes an input layer, a hidden layer and an output layer, wherein the number of the input layer is determined according to the influencing path and the risk point and its motivation, the number and structure of the hidden layer are determined according to the genetic algorithm, and the output of the output layer is the investment return risk prediction value;

[0118] Using the training set, a back propagation algorithm is applied to adjust the network weights and thresholds of the BP neural network to obtain the first model.

[0119] In some embodiments, obtaining the investment return risk probability prediction model further includes:

[0120] Calculating the prediction error of the trained BP neural network using the verification set;

[0121] The real-time impact indicator is input into the first model, and a risk assessment result is obtained based on the first model.

[0122] Specifically, based on the results of constructing the risk points and their motivation system of pumped storage power stations, a BP neural network is constructed to predict the risk points and their motivation variable levels, and the investment return risk prediction of pumped storage power stations is realized. The specific steps are as follows:

[0123] Normalizing the risk points such as historical cost data, historical new investment and their related variable data in the first relevant data;

[0124] Dividing the risk point variable data and the corresponding influencing variable data into multiple data subsets, wherein one subset of the multiple data subsets is used as a validation set, and the remaining subsets are used as training sets;

[0125] Before training the neural network, a genetic algorithm is used to optimize the network structure and initial weights;

[0126] According to the network structure and initial weights optimized by the genetic algorithm, the BP neural network is constructed, and the BP neural network includes an input layer, a hidden layer and an output layer, wherein the number of the input layer is determined according to the influencing path and the risk point and its motivation, the number and structure of the hidden layer are determined according to the genetic algorithm, and the output of the output layer is the investment return risk prediction value;

[0127] Using the training set, applying a back propagation algorithm to adjust the network weights and thresholds of the BP neural network to obtain the investment return risk probability prediction model;

[0128] Calculating the prediction error of the trained BP neural network using the verification set; and

[0129] The real-time impact index is input into the investment return risk prediction model to predict the power plant investment return risk in real time.

[0130] The neural network is characterized in that, according to the network structure and initial weights optimized by the genetic algorithm, constructing the BP neural network further comprises:

[0131] The initialized BP neural network is expressed by the following formula:

[0132]

[0133] Among them, X is the neuron input vector; n is the number of input layer nodes; Y is the neuron output vector, and m is the number of output layer nodes;

[0134] The output value H of the hidden layer is expressed by the following formula: aj :

[0135]

[0136] Wherein, p is the number of nodes in the hidden layer; ω ij is the weight between the input layer and the hidden layer; a is the threshold; f is the activation function;

[0137] The output value H of the output layer is expressed by the following formula bk:

[0138]

[0139] Among them, ω jk is the weight between the hidden layer and the output layer; b is the threshold value;

[0140] The predicted output H is expressed by the following formula bk With the expected output Y k The error between k :

[0141] e k =|Y k -H bk |;

[0142] Among them, when max(e k )>e0, the BP neural network parameters are adjusted and learning and training are performed again until all training data meet the specified error standard, where e0 is the maximum allowable error value;

[0143] Update the weight ω between the input layer and the hidden layer according to the error ij And the weight ω between the hidden layer and the output layer jk :

[0144]

[0145] ω jk =ω jk +ηH aj e k ;

[0146] Among them, η is the learning rate, which is between 0 and 1;

[0147] The threshold is corrected by the following formula:

[0148]

[0149] λ b =λ b +e k ;

[0150] Calculating the prediction error of the trained BP neural network using the verification set;

[0151] The real-time impact indicator is input into the first model, and a risk assessment result is obtained based on the first model.

[0152] In some embodiments, when the risk assessment result is obtained based on system dynamics, after receiving the first relevant data, the method includes:

[0153] Input the initial investment of the power plant and historical operation and maintenance cost parameters, and build the power plant investment, cost and pricing model based on the risk points and their related variable fitting results to calculate the capacity price benchmark value of the pumped storage power plant;

[0154] Based on the probability prediction results of the power plant risk points and related variables, combined with the driving relationship between the risk points and related variables, the first model is constructed;

[0155] The capacity electricity price and its probability distribution output by the first model are compared with the benchmark value to obtain the risk assessment result.

[0156] In some embodiments, based on the probability prediction results of the power plant risk points and related variable levels, combined with the driving relationship between the risk points and the related variables, constructing the first model includes:

[0157] Based on the investment, cost and pricing model of the pumped storage power station, combined with the probability prediction results of the risk points and their related variables, a probability function model of the corresponding variables is constructed;

[0158] The probability model of the risk point and its related variables is introduced into the corresponding formula in the investment, cost and pricing model of the pumped storage power station, and the first model is constructed through system dynamics.

[0159] Specifically, obtaining risk assessment results based on system dynamics includes the following steps:

[0160] Input parameters such as the initial investment of the power plant and historical operation and maintenance fees, and build a power plant investment, cost, and pricing model based on the fitting results of risk points and related variables to calculate the capacity electricity price benchmark value of the pumped-storage power plant; secondly, based on the probability prediction results of the risk points and related variable levels of the power plant, combined with the driving relationship between risk points and related variables, and the linkage relationship between investment, cost, and pricing, introduce system dynamics theory to simulate and analyze the probability distribution of capacity electricity price verification results; finally, compare the capacity electricity price probability distribution obtained by simulation with the capacity electricity price benchmark value to analyze the investment return risk of the pumped-storage power plant.

[0161] Since the investment cost is a known parameter and the impact of investment on cost, pricing and other stages is mainly realized through the variable of forming fixed assets, only factors such as approved investment, capital investment, and long-term loan principal are considered to construct the investment model for the construction phase.

[0162] l c = l cap +l loa ;

[0163] l c0 =α×l c ;

[0164] In the formula, l c Approve fixed asset investment for power stations;cap For capital investment; loa The principal of long-term loans; c0 is the original value of the initial fixed assets formed; α is the fixed assets formation ratio.

[0165] The operating costs of pumped storage power stations include operation and maintenance costs, depreciation, amortization, and financial expenses. The construction of their respective models is as follows:

[0166] 1) Operation and maintenance cost model

[0167] ① Material costs: refers to various raw materials, low-value consumables, auxiliary materials, etc. used in production.

[0168] O c =Q×l c ;

[0169] In the formula, O c is the material cost incurred during the operation of the pumped storage power station; Q is the installed capacity of the pumped storage power station; l c Set a quota for material costs.

[0170] ② Employee compensation: expenses incurred in accordance with relevant wage and welfare policies and employee composition stipulated by the state.

[0171] O z = l z ×R;

[0172] O l =O z ×γ;

[0173] In the formula, O z The total salary of the employees in the operation stage of the pumped storage power station; l The wages and benefits of the employees in the operation stage of the pumped storage power station; z is the wage standard for regular employees; R is the number of employees in the power station; γ is the wage and welfare rate.

[0174] ③ Repair costs: refers to the various expenses incurred for regular or irregular replacement and repair to ensure or restore the normal functions of fixed assets, which can be divided into fixed repair costs and variable repair costs.

[0175] O m =(l c ×α)×l m ;

[0176] In the formula, O m The annual maintenance cost of the pumped storage power station; c is the fixed asset investment of the power station; α is the fixed asset formation ratio; l m For repair rate.

[0177] ④ External electricity charges: refers to the cost of purchasing electricity for production needs of power plants. For pumped-storage power stations, it mainly refers to the pumping electricity charges.

[0178] O u =Q in ×p in ;

[0179] In the formula, O u Q is the electricity fee for pumping water from the pumped storage power station; in is the pumped electricity of the pumped storage power station; in For pumped water electricity prices.

[0180] ⑤ Insurance premium: refers to fixed asset insurance and other insurance. The insurance premium rate can be calculated at 0.25% of the fixed asset value.

[0181] O b =(l c ×α)×l b ;

[0182] In the formula, O b The annual insurance premium for the pumped storage power station; b For insurance premium rate.

[0183] ⑥ Water resource fee: refers to the water resource fee paid to local governments according to the prescribed calculation standards or agreements and water consumption. According to the "Notice on Adjusting the Collection Standards of Water Resource Fees for Water Used by Centrally Affiliated and Inter-provincial Hydropower Generation" (Development and Reform Price

[2014] No. 1959), water resource fees are temporarily exempted for water used in pumped storage power stations.

[0184] ⑦ Other expenses: refers to other expenses that can be included in the cost according to regulations, in addition to financial expenses such as office expenses, business entertainment expenses, water and electricity expenses, travel expenses, and transportation expenses.

[0185] O e =Q×l e ;

[0186] In the formula, O e Other annual expenses of the pumped storage power station; e Set a quota for other expenses of the power station.

[0187] 2) Depreciation

[0188] The formula for calculating the depreciation cost of a pumped storage power station is as follows.

[0189] O d =[(l c ×α)×(1-π)] / N d ;

[0190] In the formula, O dis the annual depreciation fee; π is the residual value rate; N d The depreciation period.

[0191] 3) Amortization

[0192] Pumped storage power stations generally do not have depreciation charges.

[0193] 4) Financial expenses

[0194] Financial expenses refer to various expenses incurred for raising funds, including interest expenses, exchange losses, related fees and other expenses incurred during the production and operation period. Among them, interest expenses mainly refer to the interest on construction investment loans and working capital loans during the production period.

[0195] L=L s -l loa / N l ;

[0196] O f =L×r;

[0197] In the formula, O f is the annual financial cost; L is the initial loan balance; L s N is the loan balance of the previous year; l is the repayment period; r is the annual loan interest rate.

[0198] According to the method for determining the capacity electricity price of pumped storage power stations, the pricing model of the power station is as follows:

[0199]

[0200] In the formula, CI t is the cash inflow in year t; CO t is the cash outflow in the tth year; i is the internal rate of return on capital, which is 6.5%; t is the operating year; n is the operating period, which is 40 years.

[0201]

[0202] I pow =Q out ×p out ;

[0203] In the formula, I pow is the electricity revenue in year t; I cap is the capacity electricity fee income in year t; l c40 is the original value of fixed assets at the end of the operating period; β is the residual value rate; Q out is the amount of electricity connected to the grid by the power station; p out The on-grid electricity price for power stations.

[0204]

[0205] Based on the above cash inflow and cash outflow calculation formulas, the capacity electricity price calculation formula is shown as follows.

[0206]

[0207] In the formula, p cap is the capacity price excluding tax; C cov The capacity of power station units covered.

[0208] Based on the above construction of the pumped storage power station investment, cost and pricing model, combined with the power station investment return risk points and the probability prediction results of related variables, a probability function model of the corresponding variables is constructed. The basic form is as follows:

[0209] For discrete risk points, the probability model is:

[0210] X <![CDATA[x1]]> <![CDATA[x2]]> …… <![CDATA[x n ]]> P <![CDATA[p1]]> <![CDATA[p2]]> …… <![CDATA[p n ]]>

[0211] For continuous risk points, the probability model is:

[0212]

[0213] Where f(x) is the probability density function of x.

[0214] The probability model of risk points and their related variables is introduced into the corresponding formulas in the above-mentioned pumped storage power station investment, cost and pricing model, and a simulation model is constructed through system dynamics, such as Figure 2 shown.

[0215] Finally, the capacity electricity price and its probability distribution obtained by simulation are compared with the capacity electricity price benchmark value to analyze the investment return risk of pumped storage power stations.

[0216]

[0217] In the table, y represents the simulated capacity electricity price, and y0 represents the capacity electricity price benchmark value.

[0218] In some embodiments, Figure 3 As shown, the present application provides a pumped storage power station risk assessment method based on multi-factor interaction, which is configured in the pumped storage power station risk assessment system based on multi-factor interaction provided in the above embodiment, and the method includes:

[0219] The input terminal obtains the first relevant data and sends it to the processing terminal;

[0220] The processing terminal receives the first relevant data, and acquires a preset reference value corresponding to each risk point based on the first relevant data; the first relevant data includes a plurality of risk points;

[0221] Constructing a first model of the pumped storage power station based on a path analysis method or a BP neural network or system dynamics, and obtaining a risk assessment result;

[0222] Sending the risk assessment result to a display device;

[0223] The display device generates a target image according to the risk assessment result, and displays the target image on a display of the display device.

[0224] It can be seen from the above embodiments that the present application provides a pumped storage power station risk assessment system and method based on multi-factor interaction, the method comprising an input terminal acquiring first relevant data and sending it to a processing terminal; the processing terminal receiving the first relevant data, and acquiring a preset reference value corresponding to each risk point based on the first relevant data; the first relevant data includes multiple risk points; a first model of the pumped storage power station is constructed based on a path analysis method or a BP neural network or system dynamics, and a risk assessment result is obtained; the risk assessment result is sent to a display device; the display device is configured to generate a target image according to the risk assessment result, and display the target image on the display of the display device. The present invention can not only realize the accurate prediction of multiple risks, but also divide the risks of the trading center into low, medium and high levels through the calculated comprehensive risk score, providing a scientific basis for the risk warning and management decision of the trading center.

[0225] Similar parts between the embodiments provided in this application can be referenced to each other. The specific implementation methods provided above are only a few examples under the overall concept of this application and do not constitute a limitation on the scope of protection of this application.

Claims

1. A risk assessment system for pumped storage power stations based on multi-factor interaction, characterized in that: include: An input terminal, a processing terminal and a display device; the input terminal and the display device are respectively connected to the communication interface of the processing terminal; The input terminal is configured to: obtain first relevant data and send it to the processing terminal; The processing terminal is configured to: receiving the first relevant data, and acquiring a preset reference value corresponding to each risk point based on the first relevant data; the first relevant data includes a plurality of risk points; Constructing a first model of the pumped storage power station based on a path analysis method or a BP neural network or system dynamics, and obtaining a risk assessment result; Sending the risk assessment result to the display device; The display device is configured to generate a target image according to the risk assessment result, and display the target image on a display of the display device.

2. The risk assessment system for pumped storage power station based on multi-factor interaction according to claim 1 is characterized in that: The processing terminal includes a storage unit, a calculation unit and a control unit, wherein the storage unit and the calculation unit are electrically connected to the control unit respectively; the storage unit is used to store the calculated intermediate values ​​of the evaluation system; The calculation unit is used to perform solution calculations; the control unit is used to send control instructions.

3. The risk assessment system for pumped storage power station based on multi-factor interaction according to claim 1 is characterized in that: The first relevant data includes historical data on new investments, scrapped assets, historical operation and maintenance costs, and other investment and cost risk points.

4. The risk assessment system for pumped storage power station based on multi-factor interaction according to claim 1 is characterized in that: When the risk assessment result is obtained based on the path analysis method, after receiving the first relevant data, the method includes: Generate a multi-factor variation function based on the first relevant data fitting as a risk key point benchmark value; Monitor the actual value of each risk point and calculate the difference between the actual value and the benchmark value; A path analysis method is introduced to analyze and verify the degree of influence of factors based on the difference value, to construct a first model, and to obtain a risk assessment result based on the first model.

5. The risk assessment system for pumped storage power station based on multi-factor interaction according to claim 4 is characterized in that: Generating a multi-factor variation function based on the first related data includes: When the first correlation data is linear data, selecting a univariate or multivariate linear fitting function to generate; When the first correlation data is nonlinear data, polynomial fitting, exponential fitting or other nonlinear fitting functions are selected.

6. The risk assessment system for pumped storage power station based on multi-factor interaction according to claim 3 is characterized in that: The risk assessment results based on BP neural network include: Normalize historical cost data, historical new investment and other risk points and their related variable data; Dividing the risk point variable data and the corresponding influencing variable data into multiple data subsets, wherein one subset of the multiple data subsets is used as a validation set, and the remaining subsets are used as training sets; Before training the neural network, a genetic algorithm is used to optimize the network structure and initial weights; Constructing the BP neural network, the BP neural network includes an input layer, a hidden layer and an output layer, wherein the number of the input layer is determined according to the influencing path and the risk point and its motivation, the number and structure of the hidden layer are determined according to the genetic algorithm, and the output of the output layer is the investment return risk prediction value; Using the training set, a back propagation algorithm is applied to adjust the network weights and thresholds of the BP neural network to obtain the first model.

7. The risk assessment system for pumped storage power station based on multi-factor interaction according to claim 6 is characterized in that: Obtaining the investment return risk probability prediction model also includes: Calculating the prediction error of the trained BP neural network using the verification set; The real-time impact indicator is input into the first model, and a risk assessment result is obtained based on the first model.

8. The risk assessment system for pumped storage power station based on multi-factor interaction according to claim 7 is characterized in that: When the risk assessment result is obtained based on system dynamics, after receiving the first relevant data, the method includes: Input the initial investment of the power plant and historical operation and maintenance cost parameters, and build the power plant investment, cost and pricing model based on the risk points and their related variable fitting results to calculate the capacity price benchmark value of the pumped storage power plant; Based on the probability prediction results of the power plant risk points and related variables, combined with the driving relationship between the risk points and related variables, the first model is constructed; The capacity electricity price and its probability distribution output by the first model are compared with the benchmark value to obtain the risk assessment result.

9. The risk assessment system for pumped storage power station based on multi-factor interaction according to claim 8 is characterized in that: Based on the probability prediction results of the power plant risk points and related variables, combined with the driving relationship between risk points and related variables, the first model is constructed, including: Based on the investment, cost and pricing model of the pumped storage power station, combined with the probability prediction results of the risk points and their related variables, a probability function model of the corresponding variables is constructed; The probability model of the risk point and its related variables is introduced into the corresponding formula in the investment, cost and pricing model of the pumped storage power station, and the first model is constructed through system dynamics.

10. A risk assessment method for a pumped storage power station based on multi-factor interaction, characterized in that: In the risk assessment system for a pumped storage power station based on multi-factor interaction as described in any one of claims 1 to 9, the method comprises: The input terminal obtains the first relevant data and sends it to the processing terminal; The processing terminal receives the first relevant data, and acquires a preset reference value corresponding to each risk point based on the first relevant data; the first relevant data includes a plurality of risk points; Constructing a first model of the pumped storage power station based on a path analysis method or a BP neural network or system dynamics, and obtaining a risk assessment result; Sending the risk assessment result to a display device; The display device generates a target image according to the risk assessment result, and displays the target image on a display of the display device.

Images