This paper has demonstrated the basic use of elementary neural networks for modelling and representing driver steering behaviour in path regulation control tasks. Areas of application include uses by vehicle simulation experts who need to model and represent specific instances of driver steering control behaviour, potential on-board vehicle technologies aimed at representing and tracking driver steering control behaviour over time, and use by human factors specialists interested in representing or classifying specific families of driver steering behaviour. Extension of this and related work to other technical realms is clearly possible, as the application of this specific work to other vehicle control areas such as longitudinal headway control and collision avoidance systems. A key element in this paper was the use of time delayed sensor information in order to permit the neural network to derive necessary time derivate information in its formulations of driver steering control. This technical detail, though not especially obvious, is seen as an important reason for much of the indicated success represented here. Use of time-delayed sensor information was observed to improve training time and accuracy by the neural network representations, particularly when utilizing human steering response measurements obtained from the driving simulator and test vehicle sources. Finally, it should be noted that the results seen here apply to the very specific operating conditions of the example cases. However, the variety of driving scenarios and data sources considered in the above examples would suggest that the adaptive nature of neural networks can be used effectively for representing driver steering control behaviour under a wide range of operating conditions. This could also apply to temporal variations in driver control as induced as a result of changing driving conditions or driver properties. Further research into issues of practical implementation of neural network concepts and adaptive control is anticipated in future work.
Application of elementary neural networks and preview sensors for representing driver steering control behaviour
Die Anwendung elementarer neuronaler Netzwerke und Vorerkennungssensoren zur Darstellung des Regelungsverhaltens beim Lenken von Fahrzeugen
Vehicle System Dynamics ; 25 , 1 ; 3-30
1996
28 Seiten, 16 Bilder, 35 Quellen
Article (Journal)
English
Taylor & Francis Verlag | 1996
|Automotive engineering | 1996
|Preview based Vehicle Steering Control using Neural Networks
British Library Conference Proceedings | 2013
|Preview based Vehicle Steering Control using Neural Networks
SAE Technical Papers | 2013
|