Networked control of interconnected systems with identical subsystems

The presented thesis is devoted to interconnected systems consisting of identical subsystems where the subsystems are controlled by associated local controllers. A communication network allows an information exchange among the subsystems. The arising controller structure is denoted as networked controller. This thesis proposes strategies to reduce the amount of the system information which is necessary for the design of the networked controllers (offline information reduction) and for the implementation of the designed controllers (online information reduction). The introduced strategies can be applied to interconnected systems consisting of an arbitrary (but finite) number of subsystems. The subsystem interconnections are either caused by the physical relations between the subsystems (physically interconnected systems) or have to be introduced by the controllers to cope with shared (cooperative) control goals (multi-agent systems). This thesis can be divided into three complementary parts. The first part deals with the LQR design problem for interconnected systems. Even if the isolated subsystems have simple dynamics, the overall system displays a complex behavior, if the number of the subsystems is large. A natural way to reduce the complexity of the controller design problem is to apply a decomposition method. A decomposition approach based on a state transformation is introduced in the first part which allows to design the optimal controller for an interconnected system by considering modified subsystems independently. The proposed decomposition approach can be uniformly applied to multi-agent systems and physically interconnected systems. The second part of the present thesis studies networked control of physically interconnected systems with situation dependent communication. In order to reduce the communication load, the information exchange will be only temporarily invoked among the local controllers. Based on the availability of the communicated information, the networked controller switches between two operation modes: Autonomous mode and cooperative mode. Hence, the controller design does not only include the determination of controller parameters but also the determination of the switching condition and communication topology in cooperative mode, such that a significant improvement in the system performance is obtained using the additional information received via the communication network. The third part introduces an event-based control strategy for multi-agent systems. Event-based control is a means to reduce the communication effort by invoking an information exchange among the subsystems only when the deviation between the estimated and current subsystem state exceeds an event threshold. An eventbased controller has been proposed, which mimics the continuous state-feedback controller with a desired precision. The relation between the event threshold and the approximation error is analyzed.