We found a match
Your institution may have access to this item. Find your institution then sign in to continue.
- Title
WT-ARIMA Combination Modelling for Short-term Load Forecasting.
- Authors
Jinxing Che; Huachao Zhai
- Abstract
Accurate load forecasting is of critical importance for the smart grid. However, power load forecasting accuracy is limited in the traditional ARIMA model due to the non-stationary power grid data. In this research, a combined forecasting model based on wavelet transform and ARIMA is proposed to improve the forecasting accuracy and increase the predictability of power load data. The non-stationarity of the load data is reduced, and the predictability of the load data is increased by decomposing the original data. Finally, the prediction result is a linear superposition of each subsequence's predicted value. The feasibility of the proposed comprehensive forecasting model is verified by an experiment with the actual load data in a county in Jiangxi Province. The experimental results demonstrate that the proposed WT-ARIMA model has good performance in terms of MAPE and RMSE. Compared with the traditional ARIMA model, the prediction accuracy of the WT-ARIMA model is more stable.
- Subjects
JIANGXI Sheng (China); LOAD forecasting (Electric power systems); FORECASTING; WAVELET transforms; BOX-Jenkins forecasting; ELECTRIC power distribution grids
- Publication
IAENG International Journal of Computer Science, 2022, Vol 49, Issue 2, p542
- ISSN
1819-656X
- Publication type
Article