Research on a model of system state prognosis using condition monitoring information

In this paper, we report a model to predict the condition of a monitored system based upon measured condition monitoring information to date. The true state of the monitored system is unobserved, but is assumed to be correlated with the measured condition monitoring information stochastically. The stochastic filtering technique is used to construct the relationship between the underlying true state of the monitored system and measured condition monitoring information. Not only is the model dynamic, namely, it is updated once a new piece of monitoring signal becomes measured, but also makes the most of the whole history of measured condition monitoring information rather than only the current observation which lacks practical justification. At the same time, the method of model parameter estimation is also discussed. A case example is presented to illustrate the modeling ideas