Analysis of the Effect of Atmosphere Condition on Discharge Characteristic of Air Gap and the Application of Neural Network

Abstract

The discharge voltage of air gap is an important factor to determine the level of external insulation. In this paper, in order to study the influence of atmosphere condition on the discharge voltage, an automatic discharge device has been developed to monitor the discharging experiment of sphere- sphere electrodes and record the atmosphere condition and the discharge voltage. This paper discusses the application of back-propogation neural network (BPNN) in the prediction of discharge voltage. The result indicates that it is feasible and the average relative error is 1.8%. When the BPNN based on data of one monitoring sites is used to predict the discharge voltage of other regions, the average relative error will be greater but the error is still under acceptable limits.