Research Article
Comparison of the Accuracy of Predicting Electricity Revenue by Predicting Electricity Quantity and Predicting Electricity Revenue by Historical Received Electricity Fees
@INPROCEEDINGS{10.4108/eai.8-12-2023.2344731, author={Weipeng Qi and Heng Yue and Siyang Xu and Xin Jin and Leilei Zhao}, title={Comparison of the Accuracy of Predicting Electricity Revenue by Predicting Electricity Quantity and Predicting Electricity Revenue by Historical Received Electricity Fees}, proceedings={Proceedings of the 5th Management Science Informatization and Economic Innovation Development Conference, MSIEID 2023, December 8--10, 2023, Guangzhou, China}, publisher={EAI}, proceedings_a={MSIEID}, year={2024}, month={4}, keywords={data mining analysis; winston method; monthly electricity sales forecast}, doi={10.4108/eai.8-12-2023.2344731} }- Weipeng Qi
Heng Yue
Siyang Xu
Xin Jin
Leilei Zhao
Year: 2024
Comparison of the Accuracy of Predicting Electricity Revenue by Predicting Electricity Quantity and Predicting Electricity Revenue by Historical Received Electricity Fees
MSIEID
EAI
DOI: 10.4108/eai.8-12-2023.2344731
Abstract
Through in-depth research and analysis of historical data on regional electricity sales, we have successfully revealed the potential laws and unique characteristics of its development. Based on this research, we propose an innovative monthly electricity sales forecasting method. This method first predicts the quarterly electricity sales of the target month in the quarter, and then accurately predicts the electricity sales of the target month based on the proportion to the quarter and the predicted value of the quarterly electricity sales in the quarter. At the same time, we have considered the impact of the Spring Festival period and made revisions as needed. We used this new method to predict the monthly electricity sales of a city in Jiangsu from April 2020 to December 2022. The results show that our prediction has an average relative error of 2.35%, which is significantly improved compared to the previous method. This demonstrates the excellent performance of our proposed prediction method in terms of accuracy and reliability. This innovative method provides an effective and feasible new approach for monthly electricity sales forecasting, and provides useful references for research and practice in related fields.