Using the orthogonal experiment method, dynamic energy simulation software of DeSt, and the multi-dimensional linear regressions method, the factors that influence the energy consumption in residential buildings were studied, also three multiple linear regression equations which could indicate the relationships among cooling load, heating load and the total load in a whole year were achieved. In addition, the error analysis was carried ont. By comparing the calculated results of prediction equations and simulation software, the reliability of equations was validated. Therefore, this method could be considered as a simple and available strategy to predict the energy consumption in residential buildings in Chongqing. With the reliable prediction equations, it was concluded that effects of different design methods and envelope structures on building energy consumption could be quickly and easily obtained.
energy consumption simulation
multi-dimensional linear regression
load predication equation