A Prediction Model for Electricity Consumption in the Beijing-Tianjin-Tangshan Region Based on Informer

• Published in: 2024 IEEE 3rd International Conference on Electrical Engineering, Big Data and Algorithms (EEBDA) pp. 1136 - 1141
• Main Authors: Li, Yanbin, Yu, Le, Chen, Xi, Shi, Fan, Xie, Xu, Ruan, Yangrui
• Format: Conference Proceeding
• Language: English
• Published: IEEE 27.02.2024
• Subjects: Analytical models; Clustering algorithms; DBSCAN; Dimensionality reduction; Electricity; Informer; meteorological data; PCA; Prediction algorithms; Predictive models; Time series analysis
• DOI: 10.1109/EEBDA60612.2024.10485657

Currently, the Beijing-Tianjin-Tangshan region is in an important period of transformation, with the construction and promotion of the electricity market requiring more diverse and accurate electricity services. In the prediction, climate change has a certain impact on medium to long-term changes in electricity consumption. However, due to the large amount of meteorological data and difficulties in processing, previous studies have found it difficult to consider multiple meteorological data. In this paper, the meteorological data is firstly clustered and analyzed by clustering algorithm named DBSCAN, and then the data in the same cluster is dimensionality reduced. Then, based on the dimensionality reduction data, the informer model is trained using the electricity consumption data and the meteorological data. Finally, based on the case study using the data sets from Beijing-Tianjin-Tangshan region, the prediction model for electricity consumption proposed in the paper is compared with other models to verify the effectiveness of the proposed algorithm.