ABSTRACT
To our best knowledge, almost all of the HMMbased frameworks in the literature consider only the mono-mode case, meaning that there is only one degradation mode occurring at one time. In practice, however, multiple degradation modes could exist in competition. In this paper, we propose a new HMM-based structure called MultiBranch Hidden Markov Model (MB-HMM) to deal with this problem. In order to evaluate the effectiveness of the proposed model with respect
1 INTRODUCTION
Condition-Based Maintenance (CBM) plays nowadays an important role in the maintenance of engineering systems since it has the ability to recommend in advance proper maintenance actions by basing on the information collected from condition monitoring (Jardine et al. 2006).. Within a CBM program, diagnostics and prognostics are two important aspects where the former deals with fault detection, isolation and identification when it occurs and the latter helps to assess the current health state of the system and predict the remaining time before a failure occurs. An accurate prediction of the Remaining Useful Life (RUL) could provide ample time for maintenance engineers to schedule a repair, and to acquire replacement components before they actually fail (Zhang et al. 2005). Due to the stochastic nature of the deterioration, stochastic processes have been adapted to model the deterioration processes and have shown the promising results within the CBM framework (Van Noortwijk 2009, Le Son et al. 2012, Wang and Wang 2012).
