Pilot time-varying control behavior modeling in refractory period with aircraft failures

Modeling human pilot control behavior aims to understand and describe how humans control aircrafts and devices, to provide a foundation for the study of the dynamic characteristics of the human-vehicle system. In the presence of aircraft failures, the human pilot has a control process of the refractory period, which may cause adverse aircraft-pilot couplings, and even lead to loss-of-control events. This refractory period will make the pilot emerge with time-varying and adaptive features. This paper investigates how pilot control behavior changes to adapt to the aircraft failure situation and develops a time-varying pilot model during the refractory period. Six aviation pilots performed a human-in-the-loop simulation experiment on a ground flight simulator to simulate the failure situations for a pitch-tracking task. To characterize the pilot’s time-varying response mechanism, a time-frequency-spectrum method was used to analyze the pilot control signal. Main innovations in the proposed model can be embodied in the description of the fuzziness, time-varying, and adaptation of the pilot for the failures in the refractory period. Based on fuzzy logic theory, the pilot’s judgment and identification of failures are described. The adaptation of manual control behavior to time-varying aircraft dynamics is depicted by adaptive model theory. Time-domain and time-frequency-spectrum analysis show that the simulation results of the pilot model are consistent with the human-in-the-loop experimental results. The model simulation evaluations are within the range of the experimental evaluation, which shows the rationality of the time-varying behavior model of the human pilot in a failure refractory period. The model has practical values for guiding the pilot to deal with abnormal conditions and predicting nonlinear aircraft-pilot couplings.