A comparative study on prediction of human operator violation using neural networks

This paper con tributes to the comparative prediction study by the artificial neural networks, taking into account uncertainty, of the deviated intentional behaviors of the human operators in the human-machine systems. This type of behaviors is a particular human operator violation called barrier removal. A predictive BCD model is developed by considering a multi-reference, multi-factor and multi-criteria based evaluation. As human operator´s evaluation can be uncertain, the objective of this research work is to integrate the uncertainty on the subjective judgments in the prediction of barrier removal. A comparative study is therefore implemented between the prediction without consideration of uncertainty and the prediction with consideration of uncertainty. The proposed approach is validated by a railway application in a European project named urban guided transport management system.