Research On Power System Ultra-short Term Load Forecasting Based On Machine Self-learning

Ultra-short term load forecasting means forecasting the load in an hour.It is very important to the control,the operation of the electrical power system. Improving itsaccuracy can strengthen the level of prevention and control and the efficiency of the electrical power system.The paper studied Ultra-short term the load forecasting from two aspects:data preprocessing and forecast method. Firstly, in order to select forecasting samples in Ultra-short term load forecasting, grey correlation analysis theory was used. Secondly, on the basis of learning the foundational theory of SVM, the LSSVM is introduced and the modified grid search method is advanced, so the optimizing selection with parameter of SVM supported is accomplished. Finally, the combination forecasting based on meta-learning is introduced which ensure that the weight coefficient between 0 and 1. The real power system ultra-term load forecasting proved that effectiveness of the method from the paper in the improvement of the precision of ultra-short term load forecasting.