Research on Multi-energy Coupled Power Load Prediction Based on K-means Clustering and LSTM Model

With the proposal of the goal of "carbon peak, carbon neutral",improving the utilization rate of renewable energy and ensuring the flexible use of energy system are the inevitable requirements for the development requirements of the current electricity market.Compared with the traditional energy supply mode, the integrated energy system considers the coordinated development of multi-energy coupling. In the process of power marketization, the change in users' energy consumption patterns leads to an unclear regularity of load fluctuation, while the increase of influencing factors further complicates load prediction. In this paper, the correlation between the mulit-energy coupling characteristics and influence factors is analyzed, the K-means cluster analysis of the main factors is conducted, the representative typical day is selected as the prediction sample, the LSTM model is used to predict the power load,considering the interaction between multiple energy sources. Finally, a comprehensive energy park was taken as the study case. Compare with the accuracy of the prediction data before and after this method,and caculate the proportion of each error to prove the feasibility of the method, which provides a theoretical basis for the power load prediction of multi-energy coupling.