Assessing Impacts of Vine-Copula Dependencies: Case Study of a Digital Platform for Overhead Cranes

Abstract

Usually, components in a system degrade simultaneously and, for processes such as maintenance, predictions of common failures due to degradation are needed to achieve accurate assessments for decision making. Vine copula approach used in this paper is one way of approaching dependency modelling, offering in addition, thanks to its features, flexibility when lack of data is an issue. Knowing that a multivariate vine copula approach does not have a regular structure, in this paper, we propose an algorithm to simulate correlated random numbers of a multivariate vine copula combining bivariate copulas, and the subject of study is the evaluation of the impact of the vine copula dependency structure in a risk-oriented Monte Carlo simulation model implemented in an online digital platform to support the maintenance strategies of a set of overhead cranes.