Transient stability status predictor based on relevance vector machine

In this paper, Relevance Vector Machine(RVM) based transient stability predictor for integrated power system has been proposed. Generator bus voltage measurements provided by phasor measurement units are used to train the proposed classifier. Various operating conditions are considered by simulating different types of faults like three phase to ground and line loss, on the given system. Fault clearing time is also varied from 5 cycles to 11 cycles, and loading conditions are varied from 5% to 10 % in some cases. The results obtained by RVM predictor have been compared with the results of Support Vector Machine (SVM) predictor. Proposed scheme is successfully tested on New England 39 bus system.