Nowadays contingency analysis being a part of a power system security calculations is a mandatory procedure for a short-term operational stage of the electrical power system (EPS). Nevertheless, there are still a lot of problems, concerning security calculations. For instance, recent studies have shown necessity to consider simultaneous failures of several power system elements during contingency analysis. This brings complexity and voluminosity of security calculations to the new level. On the one hand, analysis of all possible contingency combinations results in unreasonably high computing time. On the other hand, many contingency combinations, especially with very large number of simultaneously failed elements (SFE), have little impact on overall result of security estimation due to very low probability of such combinations. This paper presents novel mathematical algorithm allowing estimation of significant number of simultaneously failed elements (SNSFE). Proposed algorithm was tested on the 14-bus IEEE reliability test system. Results show adequate accuracy of the technique and also confirm the necessity to simulate simultaneous failures of several network elements.
POWER SYSTEM SECURITY ASSESSMENT AND CONTINGENCY ANALYSIS USING SUPERVISED LEARNING APPROACH
