Recently, large-scale systems (also referred to as interconnected systems in some literature) have received considerable attention due to their universal descriptions for many practical systems such as power system, space robot system, spacecraft system, multi-agent system and mechanical engineering system [1–9]. Wherein, distributed control and decentralized control as two common effective approaches are widely utilized because they can alleviate the computational burden brought by a centralized control scheme for large-scale systems. For the former, local relative exchange state information among subsystems is needed in the control protocol [7–9], whereas, the latter employs only locally available state information of each subsystem without requirement of any state information of other subsystems [1, 5]. Generally, dencentralized control is more practical because it is ubiquitous that many isolated subsystems like large-scale space robot system [6] work without too much state information exchange to finish one typical task.