Bayesian Networks for Process Monitoring

A Bayesian classifier combining statistical rules and fault signature matrix.

SPM
Bayesian Networks
Author

Achraf Cohen

Published

January 15, 2021

Overview

BNs can be developed either based on the physical and expert knowledge or using data. Both information can be complementary and could be helpful to accurately infer and determine the associations between the system’s variables M. A. Atoui, Cohen, Verron, et al. (2019).

References

Atoui, M. A., A. Cohen, P. Rauffet, and P. Berruet. 2019. “Fault Diagnosis by Bayesian Network Classifiers with a Distance Rejection Criterion.” ICINCO 2019 - Proceedings of the 16th International Conference on Informatics in Control, Automation and Robotics 1: 463–68.
Atoui, M. A., A. Cohen, S. Verron, and A. Kobi. 2019. “A Single Bayesian Network Classifier for Monitoring with Unknown Classes.” Engineering Applications of Artificial Intelligence 85: 681–90.
Atoui, M. Amine, and Achraf Cohen. 2021. “Coupling Data-Driven and Model-Based Methods to Improve Fault Diagnosis.” Computers in Industry 128 (June): 103401. https://doi.org/10.1016/j.compind.2021.103401.