Monitoring disturbances in smart grids using distributed sequential change detection

This paper considers the online disturbance detection in smart grids by using multiple sensors, which sample the power signal and communicate wirelessly to a fusion center that detects the occurrence of abnormal disturbance. Under the sequential change detection framework, we first introduce a generalized local likelihood ratio (GLLR) detector based on an autoregressive model for the disturbance. Then we propose a decentralized GLLR detector, where each sensor computes its own GLLR statistic, adaptively samples them using a level-triggered sampling scheme, and transmits the samples to the fusion center. The proposed decentralized disturbance detection scheme substantially lowers the communication overhead, while its performance is close to that of the centralized scheme.