Monthly electricity consumption prediction algorithm based on self-attention neural network

• Main Authors: CAO YOUXIA, TAO HONGCHEN, HUANG XIN, LIU MEI, NI YANYAN, ZHANG SHIKANG, HU JING, HAN DEWEI, ZHAO QIAN, XIAO PAN, CHI XINJI, GAO YUNQIANG, LI ZHI, HUANG JIN, WANG PIN, HAN HAO, TANG XU
• Format: Patent
• Language: Chinese; English
• Published: 03.05.2024
• Subjects: CALCULATING; COMPUTER SYSTEMS BASED ON SPECIFIC COMPUTATIONAL MODELS; COMPUTING; COUNTING; DATA PROCESSING SYSTEMS OR METHODS, SPECIALLY ADAPTED FORADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL, SUPERVISORYOR FORECASTING PURPOSES; PHYSICS; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE,COMMERCIAL, FINANCIAL, MANAGERIAL, SUPERVISORY OR FORECASTINGPURPOSES, NOT OTHERWISE PROVIDED FOR

The invention belongs to the technical field of electric quantity statistics, and particularly relates to a monthly electricity consumption prediction algorithm based on a self-attention neural network, and the algorithm comprises the following steps: 1, data preprocessing: carrying out the normalization processing of input data before the prediction work is carried out, and after the normalization of the input data is completed, carrying out the prediction; the monthly electricity consumption prediction model provided by the invention solves the problems of large meteorological data volume and complex characteristics when social data is considered in the electricity consumption prediction model, effectively solves the problem of difficulty in long-term time sequence prediction based on the self-attention neural network, and improves the prediction efficiency. And a stacking framework is provided, so that the prediction precision and the model generalization ability are improved. 本发明属于电量统计技术领域，具体为一种基于自注意力神经网络