HMMRel:Hidden Markov Models for Reliability and Maintenance

Reliability Analysis and Maintenance Optimization using Hidden Markov Models (HMM). The use of HMMs to model the state of a system which is not directly observable and instead certain indicators (signals) of the true situation are provided via a control system. A hidden model can provide key information about the system dependability, such as the reliability of the system and related measures. An estimation procedure is implemented based on the Baum-Welch algorithm. Classical structures such as K-out-of-N systems and Shock models are illustrated. Finally, the maintenance of the system is considered in the HMM context and two functions for new preventive maintenance strategies are considered. Maintenance efficiency is measured in terms of expected cost. Methods are described in Gamiz, Limnios, and Segovia-Garcia (2023) <doi:10.1016/j.ejor.2022.05.006>.