Residential Electricity Load Forecasting Based on Fuzzy Cluster Analysis and LSSVM with Optimization by the Fireworks Algorithm

As the construction of energy internet progresses, proportion residential electricity consumption in end-use is increasing, peak load on grid growing year year, and seasonal regional power supply tensions, mainly for consumption, have become common problems across country. Accurate forecasting can provide strong data support operation demand response incentive setting response. For accuracy stability forecasting, a model presented this paper based fuzzy cluster analysis (FC), least-squares vector machine (LSSVM), fireworks algorithm (FWA). First all, to reduce redundancy input data, it necessary dimension features. Then, FWA used optimize arguments ? ?2 LSSVM, where penalty factor denotes kernel width. Finally, method FC–FWA–LSSVM developed. Relevant from Beijing, China, are selected training tests demonstrate effectiveness proposed model. The results show that hybrid has high good versatility.