Information and communication technology are the important support technology to effective energy distribution and consumption. With the national power reform and electricity sale opened, the new requirements and goals for user demand side management and response were put forward. The efficient mining and application of consumption data should be the important basis of demand side management and response. But these data have some difficult problems as following: massive amounts, complexity of processing and analysis; In this paper, based on big data and cloud computing technology, the overall function modules of intelligent power consumption management platform were completed, and the multivariate, multi-dimensional intelligence analysis model was proposed and applied. Example as residential electricity data, the parallel outliers algorithm based on density was researched to complete the abnormal behavior and mine abnormal electricity user type; Meanwhile, Some residents in Shanghai, Beijing, Nanchang and Yinchuan were grouped, empirical research of power demand response were carried out, under the effective interaction and good incentives, the peak load can be effectively reduced; With intelligent industrial park enterprises in Gansu province as industry user example, the adaptive scheduling algorithm was proposed and applied to actual production of some enterprise with orderly electricity consumption management, the experimental result show effectively load reduction. Therefore, big data technology will provide effective support for smart power consumption in the future.