Thermal State Prediction of Transformers Based on ISSA-LSTM

Abstract

The top oil temperature is predicted to monitor the internal thermal state and operation risk of transformers. However, the existing top oil temperature prediction models usually have some disadvantages such as low accuracy, poor timeliness and difficult parameter adjustment. Therefore, a new transformer top oil temperature prediction model combining the improved sparrow search algorithm with LSTM is proposed in this paper, which can not only improve the prediction accuracy of the new model but also overcome the problems of time-consuming and laborious parameters adjustment. The advantages and a effectiveness of the proposed model are verified by simulation results using the dataset of a province.