Risk assessment of cascading failure in power systems based on uncertainty theory

Traditional risk assessment methods adopt the long-term failure model of element and severity indices to characterize the failure probability and its impact on the power system respectively. However, in the actual system, the failure probability has time-varying characteristic due to the influence of the internal system and the external environment. Severity indices are formed by weighting each severity in the system, and weight coefficients are selected based on different focus, with subjective and hypothetical premises. A new assessment method which uses the credibility measure to establish the probability of component failure and global fuzzy security index to measure the severity of the accident is proposed. N-1 contingencies are chosen as initiating event of cascading failure and risk ranking. The results of risk ranking of N-1 contingency analysis is regarded as initiating event of the next simulation layer to obtain collapse sequences. As a case study, RSTS six-bus system is simulated. The obtained results have shown that the proposed method is more accordant to actual power system operating states and it is with better rationality and validity.