Study of hydropower units fault diagnosis based on Bayesian network Noisy Or model

For the characteristics of complicated system structure and quantities of uncertain factors in hydropower units fault diagnosis, Bayesian network is introduced into the diagnosing analysis, a fault diagnosis system of hydropower units is built based on Bayesian network. Noisy Or model is introduced into Bayesian network system for overcoming the shortcomings of Bayesian network like considerable numbers of probabilities and difficulty in determining the probabilities. In this article, the network is constructed by applying experts´ knowledge in determining the probability of each node and taking the characteristic nodes as binary nodes. Then the influence of random combinations of multiple nodes on the outcome is calculated to determine the possibility of occurrence of a certain fault. The simulation indicates that the number of conditional probabilities was needed by using the model in this paper decreases from 2ⁿ to 2n; it reduces the demand for data largely and raises the speed and efficiency of hydropower units´ fault diagnosis.