Modeling of Earlier Defective State Identification Based on Condition Monitoring Information

This paper reports on a model to identify the earlier defect of a monitored system based on measured condition monitoring information to date. The true state of the monitored system is unobserved, but is assumed to be stochastically correlated with the measured condition monitoring information. We further assume that the monitored system has three states, namely good, defective and failed, and the transition of the system state follows a time dependent Markov chain. The concept of delay time is used to define a two-stage failure process and describe the transition probability between system states, and the stochastic filtering technique is used to construct the relationship between the underlying true state of the monitored system and measured condition monitoring information. At the same time, the method of model parameter estimation is also discussed. The model is simulated and the result of simulation proves the effectiveness of the model.