Fuzzy Analysis of Steady State Security for On-Line Assessment

This paper reports the application of fuzzy classification procedures to the on-line steady state security assessment. The traditional "secure" or "insecure" deterministic classification, based on limit violation indices determined from the post-contingency state, is replaced by fuzzy classification procedures. Fuzzy clustering algorithms are used to identify clusters centers h a training set, previously generated, for each considered contingency. The proposed approach is applied on a realistic model of the Hellenic Interconnected power system for one selected contingency. Performance evaluation tests were performed and the quality of the proposed classification scheme is discussed.