Multi-contingency preprocessing for security assessment using physical concepts and CQR with classifications

A new method for preprocessing the outage lists for an expert system for security assessment of power systems is presented. This method uses disturbing capability, structure and location classifications and combining qualitative reasoning (CQR) with classifications (CQRC). Because this method avoids the traditional enumerative methods such as automatic contingency selection, the number of contingencies can be reduced very sharply with heuristic common sense and rules of CQRC before numerical computing. Therefore the double contingency problem or extremely heavy cost of computing can be tackled successfully. The proposed method along with an automatic contingency analysis and classification technique can assess the classified contingencies with basic physical concepts and without network simplification, minimize the number of missing harmful contingencies efficiently and give a clear accurate output.<>