The Research On Power Grid Vulnerability Based On Probability Of Cascading Failure

Large-scale blackout which may cause a chain reaction in many important fieldssuch as energy, transportation, communication and finance becomes increasinglyfrequent in the word, leading to serious economic loss and casualties. As the basis ofthe construction of the power defense system, power grid vulnerability assessment isan extension of the power grid safety assessment and analyzes the overallvulnerabilities of networks. Therefore, in order to prevent cascading failures, it isvery significant to analyze the power grid vulnerability and vulnerable linkscomprehensively and scientifically. Aiming at this problem, the works of this paperare shown as follows.First, the network topology is modeled by using complex network theory and amethod based on PageRank algorithm for evaluating the importance of power gridnodes is proposed in this paper. The index to valuate node importance of power grid isproposed, using the principle of webpage rank in Internet. According to thecharacteristics of power grid, a modified algorithm considering the importance ofnodal load, nodal load capacity and network topology is proposed. The rationality andeffectiveness of the modified algorithm is verified through an IEEE118power grid.Then, the cascading failure probability based on probabilistic inference is putforward when the mechanism of cascading failure has been studied. Bayesian Networkmodel of cascading failure is established according to the development of cascadingfailure. Combined with the failure model of electrical components and the simulationmodel of cascading failure, the cascading failure probability is obtained through theprobabilistic reasoning of cascading failure.Finally, a method for evaluating the power grid vulnerability based on cascadingfailure probability is proposed considering the probability of cascading failure triggedby vulnerable links. Besides, the development model of cascading failure is alsosimulated to assess power grid vulnerability. The rationality and effectiveness of thismethod is verified through an IEEE39power grid.This paper mainly focued on the power grid vulnerability after cascading failurecaused by the initial failure. The method for evaluating the importance of power gridnodes based on PageRank algorithm, cascading failure probability and vulnerability ofpower grid are studied. The works indicate that the failures of critical links have an influence on power grid vulnerability. Besides strengthenning the critical links anddecreasing the cascading failure probability can all improve the power gridvulnerability. Thus, it can provide references for preventing large-scale blackout andreducing social economic loss.