Relay failure rate prediction method based on ARMA model
Through the relay failure rate prediction method based on the ARMA model, the problem of untimely monitoring of relay faults in the prior art is solved, the prediction and monitoring of relay failure rates are realized, and the reliability of relay protection devices is improved.
Patent Information
- Application Number
- CN202410859832.2
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-06-28
- Publication Date
- 2025-02-18
- Estimated Expiration
- Not applicable · inactive patent
AI Technical Summary
The existing relay operating status monitoring mainly relies on manual inspection or background alarm signals, which will affect the secondary circuit when the relay fails, which may cause refusal or erroneous movement of the protection device.
Using the relay failure rate prediction method based on the ARMA model, by collecting relay failure rate data, performing stationarity checksum differential processing, calculating the autocorrelation coefficient and partial autocorrelation coefficient, establishing the ARMA(p,q) model, and calculating the autoregression order and moving average orders to achieve the prediction of failure rate.
This method can effectively detect the operating status of the relay, prevent equipment from being damaged, improve the operating reliability of the secondary circuit, and prevent relay protection device from being refusal or erroneous.
Smart Images
Figure CN119471332A_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to the technical field of relay protection of power systems, and in particular to a relay failure rate prediction method based on an ARMA model. Background Art
[0002] Relay protection is a key technology in the power system. Its main task is to detect and respond in time when a fault or abnormal state occurs in the power system to protect the safe and stable operation of the power system. As the basic component to achieve this task, the relay ensures that the power system can respond quickly and effectively in the face of faults and abnormal conditions through detection, judgment, execution of actions and coordination with other systems, thus ensuring the safety and stability of the power system. Therefore, the operating status of the relay plays a vital role in the relay protection circuit.
[0003] The existing relay operating status is often analyzed by manual inspection or based on the alarm signal from the monitoring background. Whether it is manual inspection or monitoring background alarm, the relay is already damaged, which has affected the normal operation of the secondary circuit, and may cause the protection device to refuse to operate or malfunction. Summary of the invention
[0004] The technical problem to be solved by the present invention is to provide a relay failure rate prediction method based on the ARMA model, aiming to predict the relay failure rate time series and realize relay failure rate monitoring, analysis and early warning.
[0005] In order to solve the above technical problems, the technical solution adopted by the present invention is: A relay failure rate prediction method based on an ARMA model comprises the following steps: Step 1. Collect data; collect relay failure rate data and perform stability check. If the relay failure rate is not stable, differential processing is required, and then the relay failure rate autocorrelation coefficient is calculated. and partial autocorrelation coefficient ; Step 2, establish the model and calculate the parameters; establish the ARMA of relay failure rate ( p , q ) model and calculate the model autoregressive order p and the moving average order q ; Step 3: Prediction of relay failure rate.
[0006] The specific steps of Step 1 above are: Step 1.1. Collect the relay failure rate data in the power plant over the years according to the equipment records; Step 1.2, check the stability of the relay failure rate. If the relay failure rate is not stable, differential processing is required; Step 1.3. Calculate the autocorrelation coefficient of relay failure rate and partial autocorrelation coefficient : ; ; ; In the formula, k Indicates the interval order.
[0007] The stationarity check in the above Step 1.2 uses the time series diagram test method to determine whether the relay failure rate data is a stationary sequence: draw a time series dot graph of the relay failure rate. If each data point fluctuates around the mean and the fluctuation range does not exceed the setting, it is a stationary time series, otherwise it is a non-stationary sequence.
[0008] In the above Step 1.2, if the relay failure rate is non-stationary data, it is necessary to perform differential processing on it until the autocorrelation function value and partial correlation function value of the processed data are not significantly different from zero, and obtain x t .
[0009] The specific steps of Step 2 above are: Step 2.1, establish ( p , q )Model; ARMA p , q )The structure of the model can be described as follows: ; In the formula, p , q Respectively represent the model autoregressive order and moving average order; f and Indicates non-zero coefficients to be calculated; e t represents an independent error term; t represents a time series; x t represents the time series of relay failure rate; E represents the mean; Var represents the variance; represents the square of the fitting residual; Step 2.2, determine the order of p and q; like If the tail is truncation at q, then p = 0, and the model is expressed as MA(q); If the tail is at p, then q = 0 and the model is expressed as AR(p); and If both are tailed, the model is represented as ARMA(p,q); The model is determined by SPSS software, and the unknown coefficients in the model are calculated to obtain the prediction model of relay failure rate: .
[0010] In the above Step 3, the prediction model of the relay failure rate obtained in Step 2 is run to predict the failure rate.
[0011] The present invention provides a relay failure rate prediction method based on the ARMA model, which can detect the operating state of the relay, perform state evaluation on the relay, prevent equipment damage, improve the operating reliability of the secondary circuit, and prevent the relay protection device from refusing to operate or malfunctioning. BRIEF DESCRIPTION OF THE DRAWINGS
[0012] The present invention will be further described below in conjunction with the accompanying drawings and embodiments: Figure 1 It is a basic structural flow chart of the present invention. DETAILED DESCRIPTION
[0013] The technical solution of the present invention is described in detail below in conjunction with the accompanying drawings and embodiments.
[0014] A relay failure rate prediction method based on an ARMA model comprises the following steps: Step 1. Collect data; collect relay failure rate data and perform stability check. If the relay failure rate is not stable, differential processing is required, and then the relay failure rate autocorrelation coefficient is calculated. and partial autocorrelation coefficient ; Step 2, establish the model and calculate the parameters; establish the ARMA of relay failure rate ( p , q ) model and calculate the model autoregressive order p and the moving average order q ; Step 3: Prediction of relay failure rate.
[0015] The specific steps of Step 1 above are: Step 1.1. Collect the relay failure rate data in the power plant over the years according to the equipment records; Step 1.2, check the stability of the relay failure rate. If the relay failure rate is not stable, differential processing is required; Step 1.3. Calculate the autocorrelation coefficient of relay failure rate and partial autocorrelation coefficient : ; ; ; In the formula, k It represents the interval order. The stability is checked through the historical failure rate data. The historical failure rate data is judged. The unstable ones are differentiated to show the difference. The autocorrelation coefficient and partial autocorrelation coefficient are calculated to get the degree of connection between these data.
[0016] The stationarity check in the above Step 1.2 uses the time series diagram test method to determine whether the relay failure rate data is a stationary sequence: draw a time series dot graph of the relay failure rate. If each data point fluctuates around the mean and the fluctuation range does not exceed the setting, it is a stationary time series, otherwise it is a non-stationary sequence.
[0017] In the above Step 1.2, if the relay failure rate is non-stationary data, it is necessary to perform differential processing on it until the autocorrelation function value and partial correlation function value of the processed data are not significantly different from zero, and obtain x t .
[0018] The specific steps of Step 2 above are: Step 2.1, establish ( p , q )Model; ARMA p , q )The structure of the model can be described as follows: ; In the formula, p , q Respectively represent the model autoregressive order and moving average order; f and Indicates non-zero coefficients to be calculated; e t represents an independent error term; t represents a time series; x t represents the time series of relay failure rate; E represents the mean; Var represents the variance; represents the square of the fitting residual; Step 2.2, determine the order of p and q; like If the tail is truncation at q, then p = 0, and the model is expressed as MA(q); If the tail is at p, then q = 0 and the model is expressed as AR(p); and If both are tailed, the model is represented as ARMA(p,q); The model is determined by SPSS software, and the unknown coefficients in the model are calculated to obtain the prediction model of relay failure rate: .
[0019] By building the ARMA of the time series ( p , q ) model and determine the order of the model, we can obtain the future prediction model of the failure rate and realize future prediction judgment.
[0020] In the above Step 3, the prediction model of the relay failure rate obtained in Step 2 is run to predict the failure rate.
Claims
1. A relay failure rate prediction method based on ARMA model, characterized in that: The steps include: Step 1. Collect data; collect relay failure rate data and perform stability check. If the relay failure rate is not stable, differential processing is required, and then the relay failure rate autocorrelation coefficient is calculated. and partial autocorrelation coefficient ; Step 2, establish the model and calculate the parameters; establish the ARMA of relay failure rate ( p , q ) model and calculate the model autoregressive order p and the moving average order q ; Step 3: Prediction of relay failure rate.
2. According to a relay failure rate prediction method based on ARMA model described in claim 1, it is characterized in that: The specific steps of Step 1 are: Step 1.
1. Collect the relay failure rate data in the power plant over the years according to the equipment records; Step 1.2, check the stability of the relay failure rate. If the relay failure rate is not stable, differential processing is required; Step 1.
3. Calculate the autocorrelation coefficient of relay failure rate and partial autocorrelation coefficient : ； ； ； In the formula, k Indicates the interval order.
3. According to a relay failure rate prediction method based on ARMA model described in claim 2, it is characterized in that, The stationarity check in the Step 1.2 described above uses a time series diagram test method to determine whether the relay failure rate data is a stationary sequence: a time series dot graph of the relay failure rate is drawn. If each data point fluctuates around the mean value and the fluctuation range does not exceed the setting, it is a stationary time series, otherwise it is a non-stationary sequence.
4. According to a relay failure rate prediction method based on ARMA model described in claim 3, it is characterized in that: In the above Step 1.2, if the relay failure rate is non-stationary data, it is necessary to perform differential processing on it until the autocorrelation function value and partial correlation function value of the processed data are not significantly different from zero, and obtain x t .
5. According to a relay failure rate prediction method based on ARMA model described in claim 4, it is characterized in that: The specific steps of Step 2 are as follows: Step 2.
1. Establish ( p , q )Model; ARMA p , q )The structure of the model can be described as follows: ； In the formula, p , q Respectively represent the model autoregressive order and moving average order; φ and Indicates non-zero coefficients to be calculated; ε t represents an independent error term; t represents a time series; x t represents the time series of relay failure rate; E represents the mean; Var represents the variance; represents the square of the fitting residual; Step 2.2, determine the order of p and q; like If the tail is truncation at q, then p = 0, and the model is expressed as MA(q); If the tail is at p, then q = 0 and the model is expressed as AR(p); and If both are tailed, the model is represented as ARMA(p,q); The model is determined by SPSS software, and the unknown coefficients in the model are calculated to obtain the prediction model of relay failure rate: 。 6. According to a relay failure rate prediction method based on ARMA model described in claim 5, it is characterized in that: In the Step 3, the prediction model of the relay failure rate obtained in Step 2 is run to predict the failure rate.