Analysis And Prediction Of Power System Cascading Outages

The mechanism and prediction method of power system cascading outages are of an important meaning for network planning, operation management, control technique and market decision. The analysis and prediction of power system cascading outages can provide decision basis of dealing with emergency for operator, avoiding large blackouts.There are many factors of cascading outages, such as natural condition, equipment factors, mode of operation, load, man-made factors and so on. Some advice is given that the probability of outages is divided into the failure rate of primary device and secondary device, and the primary devices are divided into functional groups to reduce computation. It is a very complex problem to make a intact model of failure probability. There is much deficiency in the mode at present and the more effective analysis solution is necessary.Many kinds of uncertainty factors are considered in the model of failure probability. We should pay attention, since transient state hidden failures of AC-DC system can easily induce cascading outages. At last, the deficiency of probability analysis is discussed and some analysis solutions are given which are worth studying.Event tree is used to analyze cascading failures and rare event approximation is used to probability order of multiple failures, which can simplify the process of screening multiple failures. The method is applied to analysis of cascading overload. The simulation of case turned out that the method can reflect the change of failure probability following the mode of operation, and can be better to evaluate the risk of performing maintenance. The method can make further improvement for path of cascading failures of small probability and high risk, which can give some help to deeply research cascading failures. The identifying method of cascading overload is studied in the paper. The compensation method of add-on current supply is improved to save resources of computation through rational approximation, which can be called power flow distribution method. The power flow computation result of two methods respectively compares with the result of Newton-Laphson method. Their error is in the same level. It is to know that the error is mainly from reactive power through data analysis. Because the equations to depict power system usually are complex, the analytic error is got very difficultly. The different result to which order in process can be got in different case using compensation method, so the simulation method is a reasonable path.