Risk Assessment And Prevention Control Of Cascading Failures In Power System Based On Fault Chains

With the development of the electric power market and the contact of regions, electric power system is extensively distributed and faults are spreading to a large range, some fault in local grid could be extended and transmitted quickly, which may result in cascading failure and then the blackout, even the breakdown of electric power system. So the risk assessment and prevention control for cascading failures is the key to avoid the blackouts.Firstly, the models and risk assessment methods of cascading failures as well as the prevention control researches are systematically summarized in this paper. The defects of above methods on the monitoring and control for cascading failures are also analyzed, it's a need to go on further research on the practical analysis methods from theoretical source. Therefore, the view of fault chain based on security science is applied to describe and characterize cascading failures, which converts the monitoring on cascading failures into the monitoring on the links of fault chains, thereby provides a theoretical basis for prevention control of cascading failures. Moreover, according to the various influencing factors, the model based on fault chain for cascading failure is set up, and the practical generation method of fault chains based on the power path is studied.Risk assessment taking the probability and the consequence into account is introduced in power system security analysis, and three risk indicators are proposed to indicate the security risk level for prevention control decision making:system voltage instability risk, system low voltage risk and overload risk. Meanwhile, fuzzy comprehensive evaluation method is used to integrate these risk indicators and obtain the comprehensive risk and risk level, and then defensive suggestions are given by analysing the assessment results.Based on the risk indicators, a prevention control method to reduce system risk and improve system security is provided. In this method the influence of control measures on these risk indicators is quantified and the cost of control measures is also converted to the risk. Meanwhile, the control nodes are selected by the importance index of the link in fault chains, and the prevention control measures are determined to meet the requirement that the risk change is the negative maximum value by using the multi-population genetic algorithm. At last, the sum is made and the future work is point out.