Input/output selection for planar tensegrity models

A systematic method of selecting sensors and actuators is produced, efficiently selecting inputs and outputs that guarantee a desired level of performance in the H/sub /spl infin//-norm sense. The method employs an efficiently computable necessary and sufficient existence condition, using an effective search strategy. The search strategy is based on a method to generate all so-called minimal dependent sets. This method is applied to tensegrity structures. Tensegrity structures are a prime example for application of techniques that address structural problems, because they offer a lot of flexibility in choosing actuators/sensors and in choosing their mechanical structure. The selection method is demonstrated with results for a 3 stage planar tensegrity structure where all 26 tendons can be used as control device, be it actuator, sensor, or both, making up 52 devices from which to choose. In our set-up it is easy to require devices to be selected as colocated pairs, and to analyze the performance penalty associated with this restriction. Two performance criteria were explored, one is related to the dynamical stiffness of the structure, the other to vibration isolation. The optimal combinations of sensors and actuators depend on the design specifications and are really different for both performance criteria.