Affiliations: [a] School of Electrical Engineering, Dalian University of Technology, Dalian, China
| [b] State Grid Dalian Power Supply Company, Dalian, China
Correspondence:
[*]
Corresponding author. Xiongying Duan, School of Electrical Engineering, Dalian University of Technology, Dalian, China. E-mail: [email protected].
Abstract: Controlled switching strategy can reduce overvoltages due to UHV transmission line closing. However, the traditional method is difficult to determine the quantitative relationship between the closing performance of circuit breakers and the risk of failure of transmission lines. Based on the risk of failure data calculated by the statistical method, an adaptive network-based fuzzy inference system (ANFIS) model for forecasting risk of failure that occurs during the closing of transmission lines is presented. An ANFIS model is employed to map the closing performance of circuit breakers and the risk of failure of the UHV transmission line. The model is based on a system of fuzzy-rules written on the basis of previous results, the knowledge and experience. The fuzzication of input variables is carried out using fuzzy sets with the Gauss membership functions. “Grid partition” was used to generate the fuzzy inference system, and a hybrid learning rule was used to optimize the fuzzy system parameters. After the appropriate model is established through training and checking, it can forecast the risk of failure in different closing performance of circuit breakers. Then, it is easy to analyze the closing performance of circuit breakers in different risk of failure requirements.
Keywords: ANFIS, closing performance, ultra-high-voltage, controlled switching, risk of failure
