Application of ARIMA Model in Fault Diagnosis of TEP

In fault prediction, ARMA is a commonly used and important method to study time series, but it also has some problems. When time series is non-stationary, ARMA prediction effect is not accurate. Therefore, on this basis, the ARIMA model is used to transform the non-stationary time series into stationary time series by the difference method. In view of the three fault states in TEP quality, the ARIMA model is established by using Python, and a TEP quality fault diagnosis system is further established, and the quality variables are predicted. The results show that ARIMA model can predict the quality of TEP in a short time, which has the advantages of simple modeling and accurate prediction. The ARIMA model is reliable for the quality fault diagnosis. Numerical simulations show that the model can accurately describe the change of quality overtime when a fault occurred, and can make judgment and early warning of the fault in a short period of time.