Fault diagnosis for the motor drive system of urban transit based on improved Hidden Markov Model

Fault diagnosis for the motor drive system of urban rail transit could reduce the hidden danger and avoid the disaster events as far as possible. In this paper, an improved Hidden Markov Model (HMM) algorithm is proposed for fault diagnosis of motors equipment for urban rail transit. In this approach, the initial value for observation matrix B in HMM is selected based on the predictive neural network and intuitionistic fuzzy sets. Firstly, by predictive neural network the observation probability matrix B is described qualitatively based on its mathematical explanation. Then, the quartering approach is introduced to define the rules between non-membership degree and observation probability matrix B, which obtains the matrix B quantitatively. Next, the selection algorithm for matrix B is given. Finally, the experiments about the motor drive system fault diagnosis of the urban rail transit are made to prove the feasibility for the proposed algorithm.