Improved fuzzy feedback linearization and Sinswat-transformation control of inverted pendulum

This paper studies the output tracking and almost disturbance decoupling problem of nonlinear control systems with uncertainties via fuzzy logic control and feedback linearization approach. The main contribution of this study is to construct a controller, under appropriate conditions, such that the resulting closed-loop system enjoys for any initial condition and bounded tracking signal the following characteristics: input-to-state stability with respect to disturbance inputs and almost disturbance decoupling, i.e., the influence of disturbances on the L-2 norm of the output tracking error can be arbitrarily attenuated by increasing some adjustable parameters. The underlying theoretical approaches are the differential geometry approach and the composite Lyapunov approach. One example, which cannot be solved by the approach from the first paper (Marino et al., 1989) on the almost disturbance decoupling problem, is proposed in this paper to exploit the fact that the almost disturbance decoupling and the convergence rate performances are easily achieved by virtue of our approach. In order to demonstrate the practical applicability, the paper takes up the study of an inverted pendulum control system.