Coded Quickest Classification for Multiple Power Quality Events in Smart Grid

Abstract

The goal of the smart grid is to develop a more reliable, secure, and environmentally friendly power grid. Unfortunately, power quality (PQ) events are more likely to happen due to unstable renewable energy sources in smart grids. We thus focus on the quickest classification, or multihypothesis quickest change detection, which jointly detects and classifies multiple abnormal PQ events. Both the classification delay and misclassification probability are aimed to be minimized. Multiple smart meters in the grid are used, where each meter transmits its local decision to a fusion center for making final decisions. For energy saving, the capacity between each meter and the fusion center is limited to be one bit. Moreover, some meters may be faulty and misleading the final decision. To combat these faulty meters under limited link capacity, a code- based framework for quickest classification is proposed. Our contribution is twofold. First, new local decision rule based on stochastic ordering theory is proposed, which has lower complexity and competing performance compared with existing matrix Cumulative Sums (CUSUM). Second, a new fusion method based on codebook switching and minimum Hamming distance rule is developed, which can significantly lower the misclassification probability.