High-Performance Spacecraft Adaptive Attitude-Tracking Control Through Attracting-Manifold Design

Abstract

Virtually every existing adaptive attitude control solution is based on the certainty-equivalence principle, which permits the adaptive controller structure to be based upon the deterministic feedback control algorithm (controller design based on nominal system information without any inertia-parameter uncertainty) and to be used in conjunction with a suitable adaptive parameter-estimation algorithm. However, one of the main drawbacks of the certainty-equivalence-based adaptive control methodology is the arbitrary degradation in closed-loop performance due to the adaptation (parameter-estimation) process, which acts like a forcing disturbance (uncertain parameter effect) imposed onto the deterministic closed-loop control dynamics. In this paper, we significantly deviate from the classical certainty-equivalence-based adaptive control framework and develop, for the first time (to our best knowledge), a noncertainty-equivalent adaptive attitude control algorithm. This novel control design process eliminates the deleterious performance-degradation effects of the certainty-equivalence controller through the introduction of a stable attracting manifold into the adaptation process, such that the resulting closed-loop adaptive attitude control dynamics recover the deterministic (ideal) case of closed-loop attitude controller performance (i.e., no uncertain parameter effects). In addition to detailed derivations of the new controller design and a rigorous sketch of all the associated stability and attitude error convergence proofs, we present numerical simulation results that not only illustrate the various features of the new noncertainty-equivalent controller design methodology but also highlight the ensuing closed-loop-performance benefits when compared with the conventional certainty-equivalence-based adaptive control schemes.