In this paper, a methodology for diagnosis and prognosis of a system in the presence of heterogeneous information using a dynamic Bayesian network (DBN) is proposed. Due to their ability to integrate heterogeneous information - information in a variety of formats from various sources - and give a probabilistic representation of a system, DBNs provide a platform naturally suited for diagnosis, prognosis, and uncertainty quantification therein. In the proposed methodology, a DBN is first constructed via an established machine learning algorithm from heterogeneous information. The DBN is then used to track the system and diagnose faults. Uncertainty in diagnosis is quantified. Remaining useful life is then estimated and the prognosis procedure validated. The methodology is demonstrated on a cantilever beam subject to fatigue loading and faults consisting of damage at the support or a crack.
展开▼
