Abstract: The research on the preference of household appliance configuration schemes is the basis to design the effective demand side management (DSM) measures for residential consumers. In the present related research, only the appliance type configuration was considered while the different power configuration of the same household appliance was neglected. Firstly, density-based spatial clustering of applications with noise(DBSCAN) method is used to implement the power clustering of high-power electrical appliances. Then, based on the statistical information of the proportion of families with various generations, and the market share of the household appliances with different power configurations, the optimization model is established to carry out the fitting analysis on the preferences of power sizes of various household appliances. Finally, through branching and pruning process, the typical household appliance configuration schemes and their preferences are obtained. The analysis shows that the proposed method can effectively screen the typical configuration schemes of household appliances in various types of families, and the preference analysis result is in accordance with the characteristics of the families' size and living space with various generations.