Adaptive Control for Input-Constrained Linear Systems

This paper proposes a direct model reference adaptive control method for linear systems with unknown parameters in the presence of input constraints. First, we used the well-known linear quadratic regulator (LQR) technique to develop a modified reference model, which is the optimal model under input constraints. Second, a model reference adaptive controller, which tracked the modified reference model instead of the reference model, was designed to compensate for parametric uncertainties. Using Lyapunov stability theory, we proved that the modified reference model tracking error converges to zero. Simulation results demonstrate the effectiveness of the proposed controller.