Modeling for enterprise energy-consuming process based on LS-SVM and NWFE

According to the necessity for researching the optimization of enterprise energy-consuming based on model, the identification method for energy-consuming link in enterprise production process was researched. In view of the existing problems of Least Squares Support Vector Machine (LS-SVM), a modeling method based on nonparametric weighted feature extraction (NWFE) and LS-SVM was proposed. NWFE was applied to intelligent data analysis for extracting typical characteristics from the training samples, and then the data were trained to construct the energy-consuming model based on on-line LS-SVM algorithm. The simulation result showed that the presented modeling method has the advantages of shorter computing time, robust and better generalization ability.