Application of a state space modeling technique to system prognostics based on a health index for condition-based maintenance

This paper presents the application of a state space model (SSM) for prognostics of an engineering system subject to degradation. A health index (HI) is inferred from a set of sensor signals to characterize the hidden health state of the system. Bayesian state estimation and prediction formulas, on the basis of the health indices modeled by the linear regression of observed signals, are carried out to sequentially update the current health state and then predict the future health state of the system. A Sequential Monte Carlo (SMC) method is used for computation. If a failure is defined in terms of a specified level of degradation, a time-to-failure distribution can be obtained based on the predicted degradation. The method is applied to a gas turbine that is simulated via a gas turbine software package and is subject to both gradual performance deterioration and abrupt faults in service. The analysis of the case study shows that the method can provide an estimate of Remaining Useful Life (RUL) with uncertainty as well as other reliability indices of interest for operators to plan effective condition-based maintenance.

► A health index is proposed for complex systems monitored by multi-sensors. ► A state space model (SSM) is used for system level prognostics based on the health index. ► A Bayesian method is used to estimate the state and parameters of the SSM jointly. ► Both gradual and sudden changes in degradation are considered.