Temperature prediction of electrical equipment based on autoregressive integrated moving average model

In this paper, the key issue of electrical equipment temperature monitoring and prediction is studied. With the actual temperature data of a certain electrical equipment, this study investigates autoregressive integrated moving average model ARIMA (p, d, q) based on the non-stationary time series difference to describe the feasibility of equipment temperature change. A preliminary model was established using Eviews6.0, and then get the optimal prediction model using enumeration method. By means of the Matlab simulation, it is shown that the model ARIMA (5, 1, 2) can fit the temperature change trend of the equipment well and predict the temperature accurately within the acceptable range of predictive error. The ARIMA model and BP neural network model are compared with each other, and the conclusion is that the ARIMA model is more suitable for the prediction of electrical equipment temperature.