Trend prediction method of economic fixed base index of power industry based on time series

Aiming at the problem of large error in the traditional trend prediction method of economic fixed base index of power industry, a trend prediction method of economic fixed base index of power industry based on time series is designed. Firstly, wavelet transform is used to exchange the data of economic fixed base index of power industry, and the data preprocessing is completed. Then, the ARIMA model is fitted to the difference series based on the time series, and the parameters are determined according to the truncation of the autocorrelation function. On this basis, the residual sequence of ARIMA model is tested. After meeting the test conditions, the forecasting algorithm of economic fixed base index trend of power industry is deduced, and the forecasting deviation is processed to get the final forecast result of fixed base index trend. The experimental results show that the prediction error of the design prediction method is much lower than that of the experimental control group, which can solve the problem of large prediction error of traditional methods.