A Theoretical Basis for Adverse Aircraft-Pilot Coupling

Publication Date


Document Type

Conference Proceeding

Publication Title

48rd European Rotorcraft Forum, ERF 2022


Currently there is no reliable quantitative method for definitively predicting the likelihood for encountering Adverse Pilot Coupling (APC) during human-in-loop operation. This work examines APC characterized by oscillations near the: 1) rigid body dynamic mode frequency, and 2) closed loop neuromuscular (NM) mode frequency. The influence of pilot control input noise (pilot response that is not linearly correlated with the forcing function) on APC was also examined. The source of data for the work was a human-in-the-loop simulation experiment using an active sidestick where the independent variables were stick force gradient, stick sensitivity, and roll lag frequency. A specific combination of the experimental variables was observed to significantly influence the occurrence and repeatability of APC. The study developed a novel method to predict the propensity for APC that incorporates pilot noise estimation as part of pilot parameter estimation. Departing from the traditional method of using a high-frequency sum-of-sines forcing function to identify a pilot's NM frequency response, a less intrusive approach leveraging internal noise as the NM forcing function is employed instead. Preliminary results indicate that pilot internal noise originates primarily from the pilot visual equalization element. An adaptive pilot model incorporating a simple objective function and the observed relationship between noise and visual equalization and is shown to produce behavior that closely matches the experimental data. Based on APC occurrence observed in the study a new metric, Relative Margin Proximity (RMP), is proposed for assessing APC propensity. The adaptive pilot model was used to investigate vehicle configurations from an experimental data base that had been rated for APC severity. The RMP values produced by the pilot model for these configurations were consistent with their APC ratings.

Funding Sponsor

San José State University


Research Foundation