Outliers Analysis of RI Long-Term Data in Fair Weather From 1000 kV UHV AC Power Lines

Abstract

Radio interference (RI) measurement data had been recorded by establishing long-term measurement stations close to the Chinese first 1000 kV UHV ac double-circuit and single-circuit transmission line. Due to the randomness and variation of RI data in fair weather, it is necessary to detect the outliers from measurement data. Two methods have been tried for analyzing the outliers by using two samples from both 1000 kV UHV AC double-circuit and single-circuit line in this letter. With nonparametric method, the outlier detection algorithm based on β ( g, h ) distribution has been proved to be more reasonable than normality assumption method.