Risk assessment based on information entropy of cascading failure in power systems

Along with the high pace of power network interconnection development in recent years, power system operational uncertainty is increasing rapidly and system dynamic behavior is becoming more and more complicated, which makes the risk assessment of cascading failures and catastrophic events in power systems more challenging. Therefore advanced methods of high reliability to assess risks of cascading failures under various power systems operation conditions need be developed. In this paper, a risk assessment framework for cascading failures based on the information entropy principle has been developed. The developed method aims at identifying the worst case cascading development among all other possibilities to support system operators in preparing countermeasures accordingly beforehand. A case study of cascading failure analysis for IEEE 30-bus system is carried out in details. N-1 contingencies are selected as the initiating events of cascading failures. The weighting factors of all types of severity (i.e. transmission line overload, bus voltage violation, real/reactive power violation of generators) are determined by entropy principle, as compared with conventional methods which use fixed values according to expertise information with different focus. Finally, the feasibility and effectiveness of this proposed approach are demonstrated through MATLAB. The results show a high potential of this proposed method for further research and application.