Research On Bayesian Network And Its Application In Power Systems Fault Diagnosis

Bayesian Network (BN) is one of the most effective theoretical models for uncertainty knowledge expression and resoning. Based on the research of basis theory about BN, Artificial Fish-swarm Algorithm (AFSA) is applied in the parameter learning of BN composed of Noisy-Or and Noisy-And nodes for the first time, the approach is expatiated and the convergence is improved by adjusting the random move's speed. The calculating results demonstrate that this parameter learning method is feasible and preferable.Owing to the distinct performance of BN about uncertainty and the traits of power systems fault diagnosis, the fault diagnosis models are establied respectively for transmission lines, busbars and transformers, based on BN composed of Noisy-Or and Noisy-And nodes and the AFSA based parameter learning method. Diagnostic results of instance prove the effectiveness and superiority of this diagnosis method.