Codesign of communication scheduling and controller of networked control systems

Abstract

Networked control systems have received increasing attention from many researchers because of their vast potential. The insertion of communication networks in a controlled system brings network-induced defects, which are mainly caused by limited network resources. This paper proposes a codesign of periodic communication scheduling and a controller using sparsity for efficient use of the network while improving initial control performance. The effectiveness of the proposed method is verified through two numerical simulations. K E Y W O R D S control system synthesis, linear matrix inequalities | INTRODUCTIONRecently, networked control systems (NCSs) have received increasing attention from many researchers because of their vast potential [1][2][3]. NCSs have communication networks between a plant and a controller. The communication networks reduce wiring, implementation, and maintenance costs. In addition, NCSs can be applied for many applications, such as teleoperated medical field, robotics, unmanned aerial/guided vehicles, electronic control in vehicle engines, and many other industries [4][5][6][7][8][9][10][11][12].The insertion of communication networks in a controlled system brings network-induced defects such as packet dropouts, time-varying delays, and packet disorders, which are mainly caused by limited network resources. These defects can be reduced by the efficient use of communication resources.To address these issues, numerous studies have been dedicated to the improvement of control performance through the efficient usage of communication resources, which is achieved through non-uniform sampling of the control system. This form of research can be roughly classified into three types: event-triggered, self-triggered, and time-triggered approaches.The event-triggered control schemes [13][14][15] decide the type of communication based on event occurrences. Typically, such an event occurs when a considered variable in a system exceeds a certain threshold. This kind of approach requires intelligent sensors that can detect event occurrences and additional power for them. Moreover, event occurrence conditions tend to be complicated when it comes to improving control performance. As a result, the codesign of event occurrence conditions and a controller requires numerous parameters to be tuned.Self-triggered control schemes [13,16,17] decide the timing of when the next control input is updated. Generally, this kind of approach assumes that the measurement is synchronised with the updated timing of the corresponding input. Therefore, the self-triggered method can be fully implemented on the controller side. However, these prerequisites, such as synchronisation, restrict the application to the asynchronous communication of multiple sensors. In the event and the self-triggered control approaches, a maximum communication rate is guaranteed, whereas in the worst-case scenario, it will tend to be a long-term communication with the maximum rate. Hence, it is important to guarantee the average communication rate rathe...