Robust multivariable control of helicopters : from mathematical models to flight tests

This thesis describes the design and flight testing of advanced robust multivariable control laws for high performance fly-by-wire helicopters. The control laws are synthesised using H optimisation, which provides robust stability against a wide class of systems with unmodelled dynamics and parametric uncertainty. This is the first time that an H-based control system has been designed and successfully tested in both ground-based simulators and in real flight, on a fly-by-wire, variable stability helicopter.;The helicopter is a multivariable and highly nonlinear system. The dynamics vary significantly with the aircraft's orientation in the three-dimensional inertial space, the magnitude and direction of the velocity and different loading configurations. This implies a high pilot workload during operational tasks. The developed control laws provide the pilot with a means to fly the aircraft safely and effectively throughout its flight envelope.;Special attention is paid to the effects of high order rotor dynamics on the control law robustness and performance, to controller implementation issues and to the effects of aircraft configurations to the perceived handling qualities of the helicopter.;For systems that undergo large parameter variations, a novel gain scheduled methodology is proposed, which not only stabilises the linearised plants within the scheduling variable region, but also achieves H performance control objectives. This method exploits the attractive observer-based structure of the H loop shaping feedback compensators.