The purpose is to design a linear memoryless state feedback decentralized controller such that,for all admissible uncertainties and actuator failures occurring among the prespecified subset of actuators,the closed-loop system is asymptotically stable.