Approximate Feedback Linearization Control of Nonlinear Systems with Parameter Uncertainties and Input Constraints

In this paper, we investigate the feasibility of an approximate feedback linearization (AFL) approach combined with probabilistic robustness analysis and design (PRAD) for a kind of constrained nonlinear systems with parameter uncertainties. The AFL approach uses an integral observer to observe and compensate the nonlinear dynamics, uncertainties, external disturbances and saturation constraints of a system. The parameters of the controller are searched by genetic algorithms to minimize a probabilistic robustness cost function, which can directly address the designing objectives. To verify the performance and usefulness of the proposed control method, an application to a simple nonlinear control problem is performed.