부산대학교 도서관

This paper presents a Markov chain based probabilistic model to get users' stochastic activity patterns and to predict the energy consumption of a smart home. These predictions are then incorporated in our prediction and feedback based proactive energy conservation (PF-PEC) algorithm, to reduce electricity cost without compromising human comfort. The experimental results show that the proposed algorithm minimizes the total energy consumption while also ensuring standard human comfort in a smart home environment.