ON COST-SENSITIVITY CONTROLLER DESIGN METHODS FOR UNCERTAIN DYNAMIC SYSTEMS.

This paper presents a point of view for control design for flexible spacecraft with limited software capability and provides a systematic integration of the problems of the number of sensors (and their location) versus controller order. The method also includes parameter sensitivity optimization. Serious numerical burdens associated with nonlinear mathematical programming problems still remain in the ground-based design calculations, however. To alleviate some of these burdens, a suboptimal controller is shown which avoids nonlinear mathematical programming problems, although high-order Riccati equations must be solved instead. This latter scheme relies upon a cost sensitivity calculation of truncation indices to indicate which eigenmodes are to be truncated from the higher-order model to form the low-order estimator design model.

ON COST-SENSITIVITY CONTROLLER DESIGN METHODS FOR UNCERTAIN DYNAMIC SYSTEMS. Academic Article