This content is not included in
your SAE MOBILUS subscription, or you are not logged in.
Neural network models of human operator performance
Technical Paper
1995-11-0035
Sector:
Language:
English
Abstract
This paper examines the feasibility of using neural networks to represent the effects of human operators in computer models of complex man/machine systems. In the suggested approach, data from man-in-the-loop simulators are used to train the networks. The method has been tested on several typical data sets using a stand-alone prototype system, consisting of a three-layer feed-forward network with a Chemotaxis training algorithm. Successful results have been obtained, and these can be used to place constraints on the quality, quantity and type of simulator data required in future applications.