Actuation constraints in multivariable flight control systems

Mark Ronald Anderson, Purdue University

Abstract

This research deals with stability and performance robustness of non-linear multivariable feedback control systems. The systems considered include linear as well as non-linear idealizations, where non-linear models are used to represent practical limitations (limiting) in the physical devices. A robustness analysis methodology, intended for use early in the control-law synthesis, is developed to evaluate stability and performance with respect to control surface deflection and rate limits specifically. The methodology consists of three steps. First, a linear actuator signal approximation is chosen based on a priori bounds, developed herein, between the time response of the linear signal approximation and the actual actuator signal. Second, the peak magnitude of the linear signal approximation is used to define a quasi-linear representation of the deflection and rate limit saturation elements using either describing function or sector methods. Third, stability and performance robustness with respect to the quasi-linear saturation models may then be evaluated using matrix-singular-value or eigenvalue techniques. The analysis methodology is used to evaluate the stability robustness of two high-authority, multivariable flight control laws for two different pilot stick force inputs to a generic forward-swept-wing aircraft. The results are validated by time response simulations.

Degree

Ph.D.

Advisors

Schmidt, Purdue University.

Subject Area

Aerospace materials

COinS