Stability analysis for reconfigurable systems with actuator saturation

Discusses a combined analytic and simulation-based approach to assessing the stability of a control law in a system that may be subject to actuator saturation due to failures and subsequent reconfiguration. The analysis is based on linearized plant dynamics, a linearized state-feedback description of the nonlinear controller dynamics, and a nonlinear actuator model. For systems of this type, a method has previously been developed that provides less conservative estimates of the domain of attraction than other available methods. The domain of attraction estimates are used to guide simulation based stability analysis. The combined analytic and simulation based stability assessment approach is implemented in RASCLE, a software package designed to interface with an arbitrary C, C++, or FORTRAN simulation. Through the combination of analytic stability estimates and automated simulation-based analysis, RASCLE can efficiently provide information about the stability of the full nonlinear system under a wide range of conditions for the purpose of validating a reconfigurable controller.