Risk Assessment Algorithm Of Overload Leading Cascading Failures

Electric energy is the basis of modern life. The basic feature of modern power system is high voltage and long distance transmission which improves the operation efficiency, simultaneously increases the risk of power grid cascading failures. Numerous researches show that power grid blackouts are closely triggered by cascading failures. Therefore studying the cascading failures inherent mechanism has become a hot research subject in power system academic field.This dissertation begins with the background introduction and the meaning of cascading failure research. The defects of present fault chain model of power grid cascading failures had been summarized and a defined method for forecasting cascading failures was presented based on fault chain model and Fuzzy C-Means. Branches importance and correlations are calculated respectively by dynamic tree theory and Apriori method.First, the prediction index is defined based on main factors which promote cascading failures. During the prediction, in order to reduce the workload and overcome the limitation of present means, this method selected the branches with the highest value of relevance as the next failure during fault chain forecasting.Then, the dynamic fault tree model of cascading failures was established due to the similarity between fault chain model and dynamic fault tree and the set of fault chains. Moreover, probability and risk were calculated and this paper analyzed importance according to dynamic fault tree theory.Finally, more hidden information is mined from the perspective of data mining. Correlation among branches which are important to control the cascading failure in manageable proportions is analyzed based on Apriori method.This paper takes the IEEE14-bus system and IEEE39-bus system as an example, the rationality of the method proposed is verified due to the comparison with other means.