The Setup of Artifical Neural Network Model for Estimating the Insulator Pollution Degree

Based on the experiments carried out in the artificial climate chamber, a new method of predicting the insulation pollution degree coordination with the artificial neural network (ANN) is proposed. The ANN is employed to build a model that describes the relationship between the factors of temperature, humidity, power frequency leakage current, the ratio of triple-harmonic in first-harmonic and the equivalent salt deposit density (ESDD). The simulation by the experiment data is carried out. The results proved that the method could predict the pollution severity of insulators more accurately. This method is able to guide the work of cleaning insulators and provide a new way to prevent the insulator flashover