Optimal and Robust Solutions for Longitudinal Flight Control of a Canard-Configured High Performance Aircraft

Abstract

The problem of longitudinal flight control for a canard-configured combat aircraft is considered in this paper. The problem is characterized by unstable dynamics, unavailability of full state vector for the feedback, and unknown but bounded disturbances affecting the system. The solutions are developed along the lines of H-infinity Measurement-Feedback Control and the Linear Quadratic Gaussian (LQG) Control. The H-infinity control approach yields a dynamic feedback law which solves the disturbance attenuation problem via measurement feedback. This dynamic measurement feedback law consists of a combination of an estimator and a full-information controller, thanks to the separation principle for H-infinity control. The LQG approach, on the other hand, uses the simpler and well-known separation principle for optimal control and leads to an estimator/regulator pair. The controllers and the estimators are simulated in a test case to compare the performance of robust (i.e., H-infinity) and optimal (i.e., LQG) designs. The robust and optimal solutions are also simulated in another test case to verify their disturbance rejection properties against the wind gusts, which primarily perturb the angle of attack. The similarities and differences in both approaches are highlighted and the results of both design approaches are compared.