关键词: 人工智能 , 隐半Markov模型 , 快速递推算法 , 故障诊断 , 故障预测

Abstract: A hidden semi-Markov model (HSMM) is an extension of hidden Markov model (HMM), designed to remove the exponential distribution of the state durations assumed in HMM by adding ex?plicit time duration parameters. Besides HSMM powerful capability of pattern classification, it allows modeling the time duration of hidden states more reasonably, therefore, it can be used for either fault diagnosis or fault prognosis. A unified framework for both fault diagnosis and prognosis based on HSMM was presented. An efficient recursive algorithm which can reduce the computation complexity and storage of HSMM was introduced into HSMM model to facilitate the proposed HSMM-based diag?nosis and prognosis. The proposed methods were applied to fault diagnosis of helicopter gearbox bear?ings and the remaining useful life (RUL) prediction of GaAs lasers respectively. The results show the validity of the methods.

Key words: artificial intelligence , hidden semi-Markov model , efficient recursive algorithm , fault di?agnosis , fault prognosis