Combined Prediction Model of Electricity Consumption Based on Grey System and ARIMA Model Correction

In this paper, an ARIMA(p,d,q) power prediction model based on grey system correction is established. To forecast electricity consumption for five periods per month for the six months, from January 2022 to March 2023, the data mean is used to replace the outliers. Then, after the data pass the time series stationarity test, an ARIMA(p,d,q) model is introduced to make seasonal predictions of the data. After the data passed the quasi-exponential law test, the grey system prediction model was established, the results were linearly fitted, and the linear coefficient was used as the trend development factor to evaluate ARIMA(p,d,q). The model is modified to produce the final solution prediction.