Parameter Estimation in a CBM Model with MultivariateObservations

摘要

A parameter estimation problem for a condition-based maintenance model is considered. Vector-valued data iscollected at discrete time points, which gives partial information about the state of a mechanical unit thatdeteriorates in continuous time. Two types of data histories are available – histories that end with observable systemfailure and histories that end when the system is suspended from operation. Our objective is to determine optimalvalues of the model parameters that maximize the likelihood function. We show how the EM algorithm and Kalmanfilter can be used to iteratively determine the optimal parameter values. A numerical example is provided toillustrate the estimation procedure.