Case study on human reliability using artificial neural networks

This paper contributes to the analysis and the prediction by the Artificial Neural Networks, taking into account uncertainty, of the deviated intentional behaviours of the human operators in the Human-Machine Systems. This type of behaviours is a particular violation called Barrier Removal. The objective of the paper is to propose a predictive Benefit-Cost-Deficit model by considering a multi-reference, multi-factor and multi-criterion based evaluation. Human operator's evaluation can be uncertain. Uncertainty on their subjective judgements is therefore integrated in the prediction of the Barrier Removal. The proposed approach is validated through a railway application within the framework of a European project Urban Guided Transport Management System. Finally, the prediction convergence of the uncertainty-integrated model is demonstrated.