Hidden Markov Models approach used for life parameters estimations

In modern electronics and in electrical applications design is very important to be able to predict the actual product life or, at least, to be able to provide the end user with a reasonable estimate of such parameter. It is important to be able to define the availability as a key parameter because, although other performance indicators (as the mean time before failures MTBF or mean time to failure MTTF) exist, they are often misused. To study the availability of an electrical, electronic or an electromechanical system, different methods can be used. The most common one relies on memory-less Markovian state space analysis due to the fact that a little information is needed, and under simple hypothesis, it is possible to gather some outcomes on the availability of steady state value. In this paper the authors, starting from classical approach of Markov models, introduce an extension known as Hidden Markov Models approach to overcome the limits of the previous one in estimating the system availability performance over time. Such a technique can be used to improve the logistic aspects connected with optimal maintenance planning. The provided dissertation in general can be used in different contexts without losing in generality.