ABSTRACT
In most cases reliability prognosis is executed because of reliability degradation of parts of the system. It means that the reliability of a system at the start differs from reliability after some time. In literature four main groups of prognostic approaches for degradation systems are described: experience-based, model-based, knowledge-based and data-driven (Gojian et al. 2009). In practice, reliability prognosis of complex systems doesn’t only use one method: sometimes engineers apply even a few approaches. The type of reliability prognosis depends on nature of degradation. For instance, some components such as controllers almost don’t have degradation during exploitation period. They have an approximately constant failure rate. But others (mechanical components) have strong degradation of reliability parameters.
