Analysis and prediction of industrial energy consumption behavior based on big data and artificial intelligence
Qiong Wu, Hongbo Ren, Shanshan Shi, Fang Chen, Sha Wan, Qifen Li
Abstract
To promote the energy-saving transformation of industrial enterprises effectively, it is important to accurately grasp the energy consumption behavior of enterprises. In this study, the multi-type industrial enterprises in an industrial park located in Shanghai are selected for analysis. Based on the collection and processing of the measured heating (steam) data, by employing the big data analysis method, the cluster analysis is carried out from different dimensions including user difference, load fluctuation and typical daily characteristics. Following which, the multi-type load characteristic curve is obtained, and the energy consumption characteristics of the whole industrial park and different types of industrial enterprises are discussed. On this basis, the energy load forecasting model of industrial enterprises is developed based on the LSTM neural network, and the effect of introducing meteorological data on the accuracy of load forecasting is analyzed.