Identification of multiple harmonic sources in cyber-physical energy system using supervised independent component analysis

Identifying harmonic sources is essential for realizing devices operating safely and power utilizing intelligently in smart home. The traditional approaches for locating sources could not set this problem properly in smart home, as less prior knowledge of residential power network could be achieved, and harmonic sources are distributed and the amounts change dynamically. In this paper, the cyber-physical energy system is employed in identifying harmonic sources in smart home, and the supervised independent component analysis is introduced to analyze the harmonic amount of each source. A wireless power quality monitoring system is built based on cyber-physical energy system to verify the performance. The experiments demonstrate the proposed supervised independent component analysis could obtain the harmonic amount of each electric device, and the row order of the mixing matrix for estimating harmonic sources is well fixed with the order of the signals to be processed.