SHE: Smart home energy management system based on social and motion behavior cognition

Home energy management attracts more and more attention in both academic and power system. Comfort requirements of occupants are necessary to be considered, which are directly affected by the human behavior. Thus profiling the unpredictable comfort requirements of occupants is one of the challenging problems. In this paper, a human behavior cognition method is proposed to predict the users' occasional activities and estimate the comfort requirements, by monitoring the data of power load, social application and mobile sensors. The power load data is applied to study the pattern of user's operation on various appliances; the social communication is analyzed to detect users' occasional activities; and the GPS data from mobile sensors is applied to locate users and estimate their arriving or leaving time. The comfort requirements, considered as a set of variables, would be modified according to the detected human behavior. Meanwhile, the optimization is triggered to minimize the electricity cost while satisfying the updated comfort requirements. These methods are introduced into the Smart Home Energy management system (SHE) to generate personalized schedule for each user. In the experiments, a number of simulations are performed under different scenarios, which are generated randomly to simulate the human behavior. SHE presents promising potential on electricity cost saving and comfort improvement for the future smart home.