ABSTRACT
Classification of the types of accidents at an early stage of the accident in nuclear power plants is crucial for proper action selection. A plant accident can be classified by its time dependent patterns related to the principal variables. The Hidden Markov Model (HMM), a double stochastic process can apply to accident diagnosis which is spatial and temporal pattern problem. The HMM is created for each accident from a set of training data by the maximum-likelihood estimation method which uses a forward-backward algorithm and a Baum-Welch reestimation algorithm. The accident diagnosis is decided by calculating which model has the highest probability for given test data. The optimal path for each model at the given observation is found by the Viterbi algorithm, the probability of optimal path is then calculated. The system uses a left-to-right model including 6 states and 22 input variables to classify 8 types of accidents and the normal state. The simulation results show that the proposed system identifies the accident types correctly. It is also shown that the diagnosis is performed well for incomplete input observation caused by sensor fault or malfunction of certain equipment.
