An Adaptive Control Simulation Study using Pilot Handling Qualities Evaluations

Abstract

Presented herein are the results of an adaptive control simulation study using metrics and pilot handling qualities on a generic commercial aviation aircraft. For this study, seven model reference adaptive control (MRAC) technologies were identified for investigation. Each technology was tuned using a consistent methodology based on metrics and specific design requirements. Additionally, each adaptive control technology was integrated with a preexisting dynamic-inversion baseline control architecture. Scenarios for evaluation were then designed using Cooper-Harper design criteria. This paper presents the development, execution, and results from the study; additional background material is provided in the references. I. Introduction HE Integrated Resilient Aircraft Control (IRAC) project is part of the Aviation Safety Program under the Aeronautics Research Mission Directorate (ARMD) at NASA. A key focus of this project is to research the use of adaptive control technologies as a risk-mitigating tool for damaged and off-nominal aircraft. In a traditional gain-scheduled design approach, the flight controller is designed by treating the aircraft’s flight envelope as a discrete space. Controls engineers then use traditional linear control techniques to shape the handling qualities of the aircraft at each of these discrete locations. In an off-nominal or damage scenario, this design approach can break down if large model uncertainties are introduced. Even in the case of a relatively robust baseline, gain-scheduled controller, aircraft performance can be severely impacted. The goal of an adaptive flight controller is to then augment the baseline controller such that larger uncertainties can be compensated for instantaneously and without a reduction in the aircraft handling qualities. If successful, these technologies then mitigate overall risk by reducing pilot workload and providing the aircraft with improved handling qualities. The use of adaptive control technologies in manned flight has at its core a component of handling qualities and pilot controller interaction. This is evidenced by the fact that the verification and validation process (VV extremely remote combinations of occurrences are not considered. In the realm of military procurement vehicles, MIL-HDBK-1797A outlines handling qualities specifications for military aircraft. Thus, irrespective of the application, pilot based evaluations of the aircraft are an integral part of maturing any control technology to a real-world platform. Given the above, this paper presents a study to evaluate the strengths and weaknesses of several adaptive control technologies through the design and execution of a handling qualities experiment. The experiment was performed at the NASA Ames Research Center using a full-motion flight simulator and NASA test pilots. In terms of format, Section II of this paper outlines the study design. Section III then presents the simulation environment and the relevant aircraft models. Section IV very briefly reviews adaptive control, the baseline control architecture, and the adaptive control technologies considered for this study; the interested reader will find significantly greater detail on