Title: A time series-based method for predicting electricity demand in industrial parks

Authors: Yurong Pan; Chaoyong Jia

Addresses: School of Mathematics and Physics, Bengbu University, Bengbu, 233000, China ' School of Mathematics and Physics, Bengbu University, Bengbu, 233000, China

Abstract: In order to accurately predict electricity demand and improve the economy and security of the power system, a time series based method for predicting electricity demand in industrial parks is proposed. Firstly, the missing values of electricity consumption data are estimated using a seasonal exponential smoothing model. Then, the missing values are supplemented and the time series is decomposed. For each decomposed part, a suitable model is selected for fitting. For long-term trends, use univariate linear regression prediction method. For seasonal changes, choose seasonal ARIMA model for modelling. For periodic changes, use Fourier analysis method for prediction. For irregular changes, combine univariate linear regression prediction method and binary linear regression prediction method for prediction. Finally, the GARCH model is introduced to test the error sequence. The experimental results show that the proposed method improves the accuracy of the prediction model and has practical application value.

Keywords: time series; industrial parks; electricity demand forecasting; seasonal ARIMA model; trend elimination method.

DOI: 10.1504/IJETP.2025.144301

International Journal of Energy Technology and Policy, 2025 Vol.20 No.1/2, pp.95 - 109

Received: 22 Apr 2024
Accepted: 08 Jul 2024
Published online: 05 Feb 2025 *

Full-text access for editors Full-text access for subscribers Purchase this article Comment on this article