Current Electric Load Forecasting Based on WOA-VMD-LSSVM

In order to improvement of the accuracy of the short-term performance system load prediction and ensure a normal operation of daily power system, forecasting short term load model on the basis of WOA-VMD-LSSVM was proposed. To begin with, The Whale Optimisation Algorithm (WOA) is used to automatically select the best search for a optimal core parameters (K-value and penalty factor α) of the the variational modal decomposition (VMD) to get the best decomposition subsequence, reduce the following applies trend information to the prediction precision, and decompose a data by using the optimized VMD; after that, the decomposed data sets are inputted into the model separately, and each subsequence's predictions are summed to produce the final prediction. After that, decomposed groups about information is input in the pattern separately, and the forecasting scores of each of the sub-sequence summed to get the latest results of the forecast. To a certain extent, it provides a scientific decision-making basis for load management and optimal power dispatching.