A two-stage ultra short-term load forecasting model for comprehensive energy based on FCM

Abstract The dynamic scheduling requirements of a two-stage integrated energy system with single-layer variable step size are complex and diverse. When conducting ultra short-term load forecasting, it is difficult to confirm the forecasting results of different types of loads, and the degree of fitting with actual values is low. Therefore, a comprehensive energy two-stage ultra short-term load forecasting model based on FCM is proposed. In this study, we use DTW distance and Pearson distance to describe the horizontal and vertical shape similarity between load curves as initial clustering centers, set the number of cluster centers in FCM to match the type of load, iterate the membership degree of load to cluster centers and cluster center matrix using the comprehensive distance between load curves until the result conditions are met, and output the clustering results. Additionally, we calculate the distribution of predicted samples within the corresponding range of each clustering result and output two-stage ultra short-term load forecasting results. In the test results, the designed model achieved a high degree of fit between the predicted electric and hydrogen loads of the two-stage integrated energy system and the actual values, providing assistance in ensuring the smooth operation of the integrated energy system.