Time to failure estimation based on degradation model with random failure threshold

As the 'point' which the system performance degrades to failure is unpredictable or random in engineering practices, it is a challenging issue to consider the failure threshold's randomicity in fault prognostic system. This paper develops an integrated methodology for time to failure estimation based on degradation model with random failure threshold. Considering both the degradation rate and the failure threshold uncertainty in failure probability analysis, the methodology for failure time forecasting is deduced, and then the fault prognostic system developing process is given in this paper. At last, the developed fault estimation method is demonstrated by an electro-hydrostatic actuator (EHA) system simulation example in aeronautical field.