An improved Markov process for modeling uncertain multi-state degradation of equipment

To address the challenges associated with multi-state evolution and uncertainty in equipment degradation within complex systems, this study proposes a Markov modeling approach incorporating state acceleration factors. Building on the traditional Markov framework, state acceleration factors are incorporated to dynamically modify the transition rates across degradation states. This approach preserves the simplicity of the exponential distribution while effectively capturing the non-stationary characteristics of degradation rates, thereby improving the model's accuracy in representing actual equipment degradation behavior. Furthermore, the proposed degradation model is embedded into a Dynamic Bayesian Network (DBN) to enable system-level reliability evaluation, with a case study and sensitivity analysis performed to demonstrate its effectiveness and applicability.