A Short-term Load Forecasting Based On Hierarchical Analysis Of Meteorological Factors

In the trace of load forecasting link, load series clustering, hierarchical analysis of meteorological factors, and load forecasting modeling have been researched in this paper. The characteristic clustering and its analysis to power load series based on Self- Organizing Feature Map(SOM) was carried out first, and then the load series in each class were decomposed into several intrinsic modes intuitive based on empirical mode decomposition(EMD). By analyzing the correlations between the intrinsic modes and weather factors with Spearman rank correlation theory, we constructed models for each mode, the models were based on Support Vector Machine(SVM) in which the parameters are optimized by Particle Swarm Optimizer(PSO). Finally, these forecasting results of each IMF are combined to obtain final forecasting result. The simulation results of TangShan show that the method has faster speed, higher precision and greater generalization abilility.