A Study of Driver State Classification Using In-vehicle Sensors by Neural Networks

A driver support system for a safeguard against human errors, which cause traffic accidents, has created considerable interests in recent years. Identification of driver states is needed for a suitable driver support system. In this paper, we have proposed the method of classifying driver states using in-vehicle sensors, for a stable measurement with the non-extra workload, by feed-forward neural networks. To investigate the realizability of the method, the data sampling experiment was carried out in the simple circumstance. The data on each driver state are obtained by ordering a driver to drive under some condition. The simple data analysis is carried out by plotting the time series, and we can see the time series of left/right gravity is the most characteristic indicator on the simple left curved course. As a simulation result, it is confirmed that the simple neural network classification system was created by the data screening method using the shape of the neural network’s learning curve.The original publication is available at JAIST Press http://www.jaist.ac.jp/library/jaist-press/index.htmlIFSR 2005 : Proceedings of the First World Congress of the International Federation for Systems Research : The New Roles of Systems Sciences For a Knowledge-based Society : Nov. 14-17, 2159, Kobe, JapanSymposium 3, Session 7 : Intelligent Information Technology and Applications Computational Intelligence (1