Forecast of the Size of Transactions Involving Adjustable Loads in the Peaking Market Based on VMD-AdaBoost-SVM Based Trading

Abstract

Predicting the adjustable loads participating in the regulation market is a key technology to achieve optimal grid dispatch and economic operation. In this paper, we propose a Variational mode decomposition (VMD)-AdaBoost-Support Vector Machine (SVM) based method for predicting the adjustable load in the regulation market. Firstly, we introduce an index named the peaking potential values representative of the real value of the loads. which is decomposed by the variational modal decomposition for data learning. To enhance the prediction accuracy, an adaptive augmentation algorithm (AdaBoost) is used to synthesize multiple SVM weak classifiers into a single strong classifier to produce an more effective classification for predicting the peak regulation potential of adjustable loads. Simulation experimental results show that the method has higher prediction accuracy and its prediction is better than that of SVM and SVM- AdaBoost algorithms.