Contribution To The Modeling Of Industrial Processes Using Oriented Object Bayesian Networks

Résumé: The development of nowadays industries is a two-edged sword, since it facilitate the way of living and provide the goods and services for the individual daily life requirements; on the other hand, it can cause catastrophic accidents that if not well handling could cause damage to human life, assets, and the environment. To avoid falling into these losses, researchers in this Field proposed and developed several methods such as; Fault tree, Event tree, Bow tie, and Bayesian network. The main objectives of this thesis is to; (1) Present a well review on application of Bayesian network in the chemical plants and process industry within the last decade (2006-2016) and present a simple application of Bayesian network in gas facility. (2) Brie y discuss the advantages of Bayesian networks to conduct a safety and risk analysis over the well-known method Fault tree. (3) Integrate Bayesian network in the process of risk management and applied it to an LNG facility, then construct an object-oriented Bayesian network OOBN for the individual networks to abstract the model.

Publié dans la revue: