Modeling Pilot Response Delays to Beacon Collision Avoidance System Commands

摘要

This report contains the results of an analysis of pilot response delays to collision avoidance commands displayed in the cockpit of a General Aviation Trainer (GAT) simulator. Data were obtained from previous tests conducted in the Beacon Collison Avoidance/GAT simulation at the National Aviation Facilities Experimental Center (NAFEC), Atlantic City, New Jersey. Subjects were general aviation pilots with a wide range of experience. Statistical curve fitting techniques are applied to response delay data. For fixed geometries, velocities, and aircraft response rates, the separation at the point of closest approach between aircraft responding to collision avoidance system (CAS) commands is inversely related to the length of pilot delay in responding to the CAS command. In experimental simulations used to measure that collision avoidance system effectiveness, the modeling of pilot response delays should be as realistic as is practicable. The results of this study provide a more realistic model than previously available. The Gamma distribution was the best distribution applicable to the queuing processes which are conceptually similar to the process that is generating the pilot response delay times. In terms of minimum error mean square, lack of bias, and uniformity of fit, the use of the Gamma distribution was found to be superior in approximating the empirical data. The recommendation to use the Gamma distribution in modeling pilot response delays in subsequent experimentation at NAFEC is made. (Author)