Isolating switch temperature prediction based on linear neural network

In order to accurately predict the working temperature of isolation switch contact, the 110kV isolation switch for substation in Wufeng country of Zhaotong City, Yunnan province is regarded as the research object. The working temperature of isolation switch contact for passive wireless sensor acquisition is regarded as measured value, using linear neural network and BP neural network to respectively predict the temperature of isolation switch. By comparing measured value with the predicted value, the result shows that the predicted value of the linear neural network is better consistent with measured value.