CN116703249B - Reliability analysis method based on CKL wind power capacity prediction - Google Patents

Abstract

The invention discloses a reliability analysis method based on CKL wind power capacity prediction, which comprises the following steps: establishing a nested ARIMA wind speed model to obtain a prediction result of original wind speed data; establishing a quadratic Copula model wind power output probability model; building a wind power output prediction model based on CKL to obtain input data of reliability analysis; setting a reliability index of a wind power plant, and taking the reliability index as a precondition of wind power output reliability analysis; and taking a predicted result as input, and adopting a Monte Carlo simulation method to analyze the capacity reliability of the wind power predicted output. The invention can realize the reliability analysis of a period of time in the future on the basis of the original reliability analysis, and solves the problems of insufficient stability, low precision and the like in the prior art.

Description

A credibility analysis method based on CKL wind power capacity prediction

Technical field

The present invention relates to the credibility analysis of wind power output capacity, especially a credibility analysis method based on CKL wind power capacity prediction.

Background technique

With the exhaustion of traditional fossil fuels, the two major problems of energy shortage and environmental pollution are becoming increasingly serious, and the transformation of the energy structure is imminent. The proposal of the "double carbon" plan has greatly promoted the adjustment of my country's energy structure, so the use of renewable and clean energy such as wind power and photovoltaics has become the main source of energy.

Due to the certain intermittency and randomness of wind power generation, it is usually treated as an uncontrollable power source. Even if wind power makes a certain contribution after being integrated into the power grid, wind power still cannot replace a certain amount of traditional power generation and can only be used as a supplementary energy source. Therefore, wind power is considered an unstable power generation method.

Contents of the invention

Purpose of the invention: The purpose of the invention is to provide a credibility analysis method based on CKL wind power capacity prediction, thereby solving the problems of insufficient stability and low accuracy in the existing technology, and enabling it to have a prediction function.

Technical solution: A credibility analysis method based on CKL wind power capacity prediction according to the present invention includes the following steps:

(1) Establish a nested ARIMA wind speed model and obtain the prediction results of the original wind speed data;

The nested ARIMA wind speed model in step (1) is specifically:

;

In the formula, p , d , q are the end parameters of the ARIMA model, expressed in the form of ARIMA( p , d , q );

;

In the formula, is a time lag variable, /> for all/> This method simulates the growth rate distribution by considering the Gaussian process that generates volatility during the growth rate process, and the process is a log-normal distribution.

(2) Establish a quadratic Copula model wind power output probability model, that is, the QC wind power output probability model.

;

In the formula, x and y are the relevant parameters of the two wind farms respectively; is the QC function correlation coefficient; when the wind speed correlation is established, the time series of the wind speed that can be determined is {t ₁ , t ₂ , t ₃ ,..., t _n }, and the wind farm edge distribution function is established based on the QC function:

;

In addition, based on considering the time series, wind speed is a discrete variable, so it is necessary to construct a joint distribution function of discrete variables:

;

In the formula, A and B are the limit states of the two electric fields respectively;

Establish wind power output model:

;

In the formula, V represents the rotation speed of the fan at time t ; V _i represents the cut-in wind speed; V _a represents the standard wind speed; P _a represents the rated output power of the fan;

Combine the marginal distribution and Copula function to establish a Copula-based wind power output model:

;

In the formula, L is the discrete coefficient; v(t) is the wind speed function that changes with time; P _N is the rated output.

(3) Establish a wind power output prediction model based on CKL and obtain input data for credibility analysis;

The CKL-based wind power output prediction model in step (3) is specifically:

;

;

In the formula, IMF _m is the m-th mode function; R _M is the M -th residual sequence;

;

;

;

In the formula, x is an environmental variable; K is an environmental change kernelization matrix; is the contribution rate;/> is the cumulative contribution rate; P is the dimensionally reduced environmental variable matrix after dimensionality reduction.

(4) Set wind farm reliability indicators as a prerequisite for wind power output credibility analysis.

The wind farm reliability indicators described in step (4) are specifically:

Loss of power probability (LOLP):

;

In the formula, t _a is the time when the power failure occurs; L is the set of power failure states; T is the total system operation time;

Loss of power time expectation (LOLE):

;

Expected loss of energy capacity (EENS):

;

In the formula, Pi is _the average power loss capacity per year.

(5) Using the prediction results as input, Monte Carlo simulation method is used to analyze the capacity credibility of the wind power forecast output. The specific formula of the credibility analysis is:

;

;

In the formula, R is the reliability of the wind power system above the standard reliability R ₀ at any time, P _W is the capacity of wind turbines to effectively replace conventional units, C _org system initial capacity, P _credit is the wind power capacity effectively connected to the system , P _wind is the total wind power capacity.

A computer storage medium on which a computer program is stored. When the computer program is executed by a processor, the above-mentioned credibility analysis method based on CKL wind power capacity prediction is implemented.

A computer device, including a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, it implements the above-mentioned credible wind power capacity prediction based on CKL. degree analysis method.

Beneficial effects: Compared with the existing technology, the present invention has the following advantages:

1. The present invention proposes a QC wind power output probability model in Copula for the correlation between wind farms, which can well describe the relationship between wind farms.

2. A CKL prediction model was constructed. Compared with a single LSTM, its RMSE and MAE were reduced by 13.3% and 19.1% respectively, and R2 was increased by 2.15%. Compared with other common methods, CKL has higher accuracy.

3. On the basis of predicted capacity, a credibility analysis is proposed, which enables the credibility assessment of wind farms to have a predictive function and can realize the credibility analysis at time t+1; and it is analyzed that energy storage equipment can improve the efficiency of wind power. The credibility of the field.

Description of the drawings

Figure 1 is a flow chart of the CKL wind power output prediction steps.

Figure 2 is a step flow chart of a credibility analysis method based on CKL wind power capacity prediction according to the present invention.

Figure 3 is the IEEE RTS-79 system topology diagram.

Detailed ways

The technical solution of the present invention will be further described below with reference to the accompanying drawings.

Figure 1 is a flow chart of the steps in the prediction part of a credibility analysis method based on CKL wind power capacity prediction provided by the present invention, which includes the following steps:

(1) Establish a nested ARIMA wind speed model and obtain the prediction results of the original wind speed data.

(2) Based on the above steps, establish a quadratic Copula model wind power output probability model, that is, the QC wind power output probability model.

(3) Establish a wind power output prediction model based on CKL and obtain input data for credibility analysis.

In order to verify the high accuracy of the CLK model of the present invention, long and short-term time prediction (LSTM), modal decomposition and long and short-term time prediction combined (EMD-LSTM) are used to compare it. It can be seen from Table 1 that the CLK model The RMSE of the state value is smaller than LSTM and EMD-LSTM, and the fluctuation is small.

Table 1 Evaluation indicators of each model

As shown in Figure 2 and Figure 3, this embodiment runs on the IEEE RTS-97 system.

(4) Set wind farm reliability indicators as a prerequisite for wind power output credibility analysis.

(5) Using the prediction results as input, Monte Carlo simulation method is used to analyze the capacity credibility of the wind power forecast output.

In order to verify the diversity of credibility analysis based on the CKL prediction model of the present invention, a comparison is made before and after adding energy storage equipment (BESS). From Table 2, it can be seen that the predicted output of the system after BESS is significantly increased, and the prediction is credible. The degree also increases.

Table 2 System reliability indicators and trusted capacity

Claims (7)

1. A credibility analysis method based on CKL wind power capacity prediction, which is characterized by including the following steps: (1) Establish a nested ARIMA wind speed model and obtain the prediction results of the original wind speed data; (2) Based on the prediction results of wind speed data, establish a quadratic Copula model wind power output probability model, that is, the QC wind power output probability model; (3) Establish the CKL-based wind power output prediction model CEEMD-KPCA-LSTM model to obtain input data for credibility analysis; the CKL-based wind power output prediction model in step (3) is specifically: ; ; In the formula, IMF _m is the m-th mode function; R _M is the M -th residual sequence; ; ; ; In the formula, x is an environmental variable; K is an environmental change kernelization matrix; is the contribution rate;/> is the cumulative contribution rate; P is the dimensionally reduced environmental variable matrix after dimensionality reduction; (4) Set wind farm reliability indicators as a prerequisite for wind power output credibility analysis; (5) Taking the prediction results obtained from the wind power output prediction model CEEMD-KPCA-LSTM model in step (3) as input, the Monte Carlo simulation method is used to analyze the capacity credibility of the wind power prediction output. 2. A credibility analysis method based on CKL wind power capacity prediction according to claim 1, characterized in that the nested ARIMA wind speed model in step (1) is specifically: ; In the formula, p , d , q are the end parameters of the ARIMA model, expressed in the form of ARIMA( p , d , q ); ; In the formula, is a time lag variable, /> for all/> This method simulates the growth rate distribution by considering the Gaussian process that generates volatility during the growth rate process, and the process is a log-normal distribution. 3. A credibility analysis method based on CKL wind power capacity prediction according to claim 1, characterized in that the QC wind power output probability model described in step (2) is specifically: ; In the formula, x and y are the relevant parameters of the two wind farms respectively; is the QC function correlation coefficient; when the wind speed correlation is established, the time series of the wind speed that can be determined is {t ₁ , t ₂ , t ₃ ,..., t _n }, and the wind farm edge distribution function is established based on the QC function: ; In addition, based on considering the time series, wind speed is a discrete variable, so it is necessary to construct a joint distribution function of discrete variables: ; In the formula, A and B are the limit states of the two electric fields respectively; Establish wind power output model: ; In the formula, V represents the rotation speed of the fan at time t ; V _i represents the cut-in wind speed; V _a represents the standard wind speed; P _a represents the rated output power of the fan; Combine the marginal distribution and Copula function to establish a Copula-based wind power output model: ; In the formula, L is the discrete coefficient; v(t) is the wind speed function that changes with time; P _N is the rated output. 4. A credibility analysis method based on CKL wind power capacity prediction according to claim 1, characterized in that the wind farm reliability index described in step (4) is specifically: Power loss probability LOLP: ; In the formula, t _a is the time when the power failure occurs; L is the set of power failure states; T is the total operating time of the system; Power loss time expectation LOLE: ; Loss capacity expectation EENS: ; In the formula, Pi is _the average power loss capacity per year. 5. A credibility analysis method based on CKL wind power capacity prediction according to claim 1, characterized in that the formula of the credibility analysis in step (5) is specifically: ; ; In the formula, R is the reliability of the wind power system above the standard reliability R ₀ at any time, P _W is the capacity of wind turbines to effectively replace conventional units, C _org system initial capacity, P _credit is the wind power capacity effectively connected to the system , P _wind is the total wind power capacity. 6. A computer storage medium with a computer program stored thereon, characterized in that, when executed by a processor, the computer program implements a feasible method based on CKL wind power capacity prediction as described in any one of claims 1-5. Reliability analysis method. 7. A computer device, comprising a memory, a processor and a computer program stored in the memory and executable on the processor, characterized in that when the processor executes the computer program, it implements claims 1-5 A credibility analysis method based on CKL wind power capacity prediction according to any one of the above.