A useful technology for PHM with signal simulation

Prognostics and health management (PHM) is a technology which seeks to reduce/eliminate inspections and time-based maintenance through accurate monitoring, incipient fault detection, and prediction of impending faults. The theories and the results of PHM technology are still in the baby stage, and the engineering applications of PHM technology need to be verified and validated, which requires a large number of simulations and tests to make it matures. An important factor restricting the development of PHM technology is the lack of system failure data. This paper presents an identically distributed resampling technology based on a small number of failure data. Namely on the basis of failure mode database, combining with statistical principles, the approach uses sampling method such as the bootstrap to achieve the RUL(Remaining Useful Life) data simulation with identically distribution for higher-quality data. The experiment with PHM08 data shows that this method can more accurately estimate the life of the engine unit. Finally the development process of a resampling module based on a small number of failure data is introduced.