Energy-saving feature extraction method for urban buildings with near-zero energy-consuming based on SVR

摘要

In order to solve the problem of poor fitting performance of traditional energy saving feature extraction method in urban buildings with near zero energy consumption, an energy saving feature extraction method based on SVR is proposed. The data are recovered and processed by means of mean value substitution method, and the correlation order of data parameters is realised through grey correlation analysis. Based on the feature weighting theory, the energy saving data of near zero energy saving buildings are cluster analysed. The main component analysis method is used to deal with feature extraction data, reduce the size of feature extraction data, and use SVR to achieve the extraction of energy saving characteristics of nearly zero energy consumption buildings in cities. The experimental results show that the method is always higher than other methods, with a maximum of 88%. The results show that the method is effective in feature extraction.