Transient Stability Analysis of Power System Based on Bayesian Networks and Main Electrical Wiring

摘要

In order to deal with the uncertainties of power system better and overcome the shortcomings of other artificial intelligence methods, a new method based on Bayesian networks and main electrical wiring was proposed. Reliability analysis methods were adopted such as depth-first search (DFS) and matrix method. Multi-state components were introduced to represent the main electrical wiring. All contingency states were obtained by minimal cut sets. Markov chain Monte Carlo (MCMC) program of approximate inference algorithm was then applied. Vulnerability was used as index to denote the weights of some vectors and was updated in real time. The example of 3/2 breakers scheme of power plant testified the feasibility of this model. It could effectively transform uncertainties into probabilities and achieve ideal results.