Robust control of input limited smart structural systems

Integration of controllers with smart structural systems require the controllers to consume less power and to be small in hardware size. These requirements pose as limits on the control input and the order of the controllers. Use of reduced order model of the plant in the controller design can cause spill over problems in the closed-loop system due to possible excitation of the unmodeled dynamics. In this paper, we present a method to design output feedback robust controllers for smart structures in the presence of control input limits considering unmodeled dynamics as additive uncertainty in the design. The performance requirements for the design are specified as regional pole placement constraints on the closed-loop poles. The controller design problem requires the maximization of damping ratio in the presence of additive uncertainty and control input limits. The resulting optimization problem for the controller design is formulated as a generalized eigenvalue problem involving linear matrix inequality (LMI) constraints. The proposed controller is designed and implemented on a multiinput-multioutput 3-mass smart structural test article. The tradeoffs involved in the controller design are analyzed and the performance and robustness specifications are verified experimentally.