Abstract
A study has been completed at LaRC which applies a generic model of an operator, the Human Operator Simulator (HOS), to the representation of a pilot flying an advanced transport airplane. The model has been demonstrated to generate results which closely match those obtained in fixed based, hardware oriented, piloted simulation studies. The model generates lookpoint predictions as could be measured with an oculometer system as well as time histories of state variables of the airplane and pilot control inputs, and provides a means of estimating pilot workload.
The model incorporates a Kalman filter type estimator to represent the pilot's knowledge of the airplane's state, and his sequential process of updating this knowledge by absorbing information from different display sources during the course of a flight.
This presentation provides a brief description of the HOS based pilot model and presents in some detail, implementation in the model of the algorithm to represent the pilot's knowledge of the state. The algorithm used allows representation of the effects of reading a single instrument at a given time. Results of some preliminary validation and demonstration steps are presented. Also, elements of a plausible pilot (or operator) scanning model, which is a spin-off of the estimation process, are discussed.
Get full access to this article
View all access options for this article.