Hidden markov model based on signals from blocks of semi-markov system’s elements and its application for dynamics analysis energy systems

Abstract

For modeling systems for various purposes, in particular, energy systems, semi-Markov processes are often used. During the functioning of the system for which the semi-Markov model is built, it is not always possible to get all the information contained in the status codes when changing its states, but you can only get the signal (information) in which block of system elements the state changed (failure, renewal, etc.). In this case, the states of the semi-Markov model can be considered hidden (unobservable). There are problems of analyzing the dynamics, predicting the states of the elements of the simulated system based on the received vector of signals from blocks of system elements. To solve these problems, the apparatus of the theory of hidden Markov models can be used. The paper considers the possibilities of applying this approach by the example of independent renewal processes superposition.