STATISTICAL MODELS OF PILOT PERCEPTION AND CONTROL AS INDICATORS SENSITIVE TO STRESS

Abstract

This chapter discusses a unified effort of analysis, simulation, and experimental study to model statistical aspects of pilot perception and control during manual control tracking. The sequential (optimal controller) model is used as the basis for this study. Statistical descriptions given by the sequential model are explored and extended. The sequential (optimal controller) model, in its general form, is denoted a sequential input adaptive system. A number of assumptions are made in the sequential model. In the sequential model, the optimal response is determined by the controller's strategy as given in a cost function. In the analysis component of this research, finite time prediction of target motion was examined and also a perturbation analysis was made. Finite time prediction, rather than infinite time prediction, most likely conforms to prediction by the human controller. A single-axis manual control tracking experiment was conducted to demonstrate how the stochastic model is applied to a man–machine tracking system. The perception and control variables in the sequential model were assumed to have modified Gaussian distributions with means and variances identified from the experimental data. The simulated tracking response, with random perception and control effects, was in general agreement with the experimental data.