Risk Analysis Of Regional Power System Cascading Blackouts

In recent years, large blackouts happened frequently all over the world, causing thegreat losses to global society and economy. These large blackouts were almost causedby cascading failure, so the research on the mechanism and risk analysis of powersystem cascading failure is very important to the improvement of power systemreliability.Based on the summary of the research status of cascading failure, the thesisanalyzes the risk of power system cascading failure from the perspectives of risk index,historical outage data based probability distribution of cascading line failures in aregional power grid and sensitivity analysis of the indexes respectively.Firstly, on the basis of Self-organized criticality theory and complex networktheory, the thesis sets up a risk analysis index system of cascading failure from theperspectives of traditional risk analysis index, probability distribution function of faultamount and System equilibrium degree respectively. Secondly, a practical validity ofbranching process model for power system risk analysis based on the transmission lineoutages recorded over14years in a regional power grid is given. The thesis improvesthe conventional branching process model and proposes a method to improve theaccuracy of estimating the probability distribution of fault amount. The thesis usestraditional probabilistic model, conventional branching process model and improvedbranching process model to estimate the probability distribution of fault amountrespectively, and then verifies the applicability of branching process model by goodnessof fit test and error analysis. Results shows that the empirical distribution of the totalnumber of line outages is approximated well by the branching process model. For thesame confidential level, the estimation of the probability distribution of the largercascades by the branching process model requires significantly fewer recorded outagedata than empirical estimation by a factor of10-1. The simulation result of improvedmodel indicates that the improved branching process model is superior to the traditionalmodel. At last, the thesis verifies the applicability of using the index system to analyzethe risk of cascading failure by studying the effect on the risk indexes when the runningstate of the system is changed. Also, the thesis gives a preliminary analysis of thecascading failure early warning that proposed in this thesis.