Input-output selection for planar tensegrity models

The input-output selection approach followed in this briefuses a rigorous and systematic procedure,efficiently selecting actuators and/or sensors that guaranteea desired level of performance, embedded in a heuristic.The procedure generates all so-called minimaldependent sets and uses a closed loop criterion.The heuristic is a divide-and-conquer one.This approach is applied to controlled tensegrity structures,using as criterion efficiently computableconditions for the existence of a stabilizing $\Hinf$-controller achievinga desired level of performance.Structural systems, like controlled tensegrities,are a prime example for application of techniquesthat address system design issues,because they present opportunitiesin choosing actuators/sensors andin choosing their mechanical structure.Results for a three-unit planar tensegrity structure,where all 26 tendons can be used as actuator or sensor devices,making up 52 devices from which to choose,demonstrate the approach.Two design specifications were explored, one is related to the dynamicalstiffness of the structure, the other to vibration isolation.The feasible sets of actuators and sensors depend on thespecifications and really differ for both, but are mostlycomposed of much less than 52 devices.