Modelling complex large scale systems using object oriented Bayesian networks (OOBN)

Abstract

The aim of this communication is to present a new way of how to structure modelling process of complex and large scale systems by object oriented Bayesian network (OOBN) for risk assessment and management purpose. In the first stage, we extend OOBN by presenting a new definition that introduces some flexibility, in a second stage, dynamic Bayesian networks (DBN) described by OOBN method are presented, that leads to a framework that we refer to as Dynamic Objet Oriented Bayesian Network (DOOBN). A demonstration in the domain of risk assessment of flash floods effect on the infrastructures inoperability is considered to show potential applicability of the extended OOBN.