作者: Youda Liu , Xue Wang , Yanchi Liu , Sujin Cui
DOI: 10.3390/S16081316
关键词:
摘要: Cyber-physical energy systems provide a networked solution for safety, reliability and efficiency problems in smart grids. On the demand side, secure trustworthy supply requires real-time supervising online power quality assessing. Harmonics measurement is necessary evaluation. However, under large-scale distributed metering architecture, harmonic faces out-of-sequence (OOSM) problem, which result of latencies sensing or communication process brings deviations data fusion. This paper depicts network asynchronous analysis exploits nonlinear autoregressive model with exogenous inputs (NARX) to reorder measuring data. The NARX gets characteristics electrical harmonics from practical rather than kinematic equations. Thus, data-aware approximates behavior parameter improves retrodiction accuracy. Theoretical demonstrates that method maintains reasonable consumption computing resources. Experiments on testbed cyber-physical system are implemented, accuracy adopted evaluate mechanism network. Results demonstrate an improvement precision validate systems.