Tirza Routtenberg

In this paper, the problem of online change-point detection of voltage imbalance in a three-phase power system using phasor measurement unit (PMU) data is considered within a sequential hypothesis-testing framework. A general model for the positive-sequence data from a PMU measurement at the time domain and off-nominal frequencies is presented. The new formulation, which assumes an additional Gaussian noise, enables fast online detection of imbalance. Closed-form expressions of the cumulative sum (CUSUM) and generalized likelihood ratio (GLR) tests are developed for detection of imbalances. The performance of the change-point detection procedures is evaluated using the average-run-length and the expected detection delay. Numerical simulations show that the proposed method can be used for enhanced situational awareness in future grid management systems and demonstrate the ability to inform strategies for advancing grid capabilities by using change-point detection methods.