Research on reliability model of relay protection device based on GM-SVR

Abstract

In order to solve the problem of low reliability parameter estimation accuracy of relay protection devices in intelligent substations under the condition of small samples, a parameter estimation method of three-parameter Weibull distribution based on GM-SVR is proposed. First of all, in view of the high reliability of the protection device at the initial stage of operation, a three-parameter Weibull distribution model is introduced. Secondly, the SVR is introduced on the basis of the GM(1,1) model, and a high-precision parameter estimation model of GM-SVR Weibull distribution is established by making use of the high reliability of the GM(1,1) model for position parameter estimation under the condition of small samples and the advantage of SVR method in small sample data processing. In order to solve the problem that the traditional model ignores the influence of equipment aging and maintenance, the equivalent fallback time and maintenance improvement coefficient are introduced to improve the time-varying model. Through the calculation of an example and comparison with other methods, it is proved that the GM-SVR method has higher accuracy of model parameter estimation, and the model considering the influence of maintenance can more effectively study the reliability of relay protection devices.