Electric Load Forecasting Based on CEEMDAN and LSSVM Optimized by Cuckoo Search Algorithm

In the paper, short-term electric load forecasting (STLF) by optimization method is studied. First, the complete ensemble empirical mode decomposition with adaptive noise (CEEMDAN) is used to decompose the short-term electric load data into some intrinsic mode functions (IMF): high frequency sequence (HF), the low frequency sequence (LF) and the trend term (T). Then, the least squares support vector machine model (LSSVM) optimized by the cuckoo search algorithm (CS) is utilized to predict electric load. Moreover, the final prediction value is obtained by the add integration method. Last but not least, the five benchmark models are added as the comparison model. The results show that the hybrid integrated model has higher horizontal precision and direction accuracy than other benchmark models. The results demonstrate that the accuracy of the model with decomposition method is higher than that without decomposition method.