A HIDDEN MARKOV DETERIORATION MODEL WITH MEASUREMENT ERRORS

Abstract

The Markov chain models have been applied in many deterioration forecasting practices. The inspection data may include measurement errors, categorized by random errors and system one by the so-called representation matters. In this paper, the deterioration processes with measurement errors are formulated as a hidden Markov chain model. The Bayesian estimation of the hidden Markov chain models can be made by use of Markov Chain Monte Carlo simulation technique. The applicability of the methodology presented in this paper is examined against the real world data concerning the national road pavement.