Self-organizing control of physically interconnected systems for disturbance attenuation

This paper presents a self-organizing networked controller for the disturbance attenuation in physically interconnected systems. To avoid large control errors, the local controllers send information to the other local controllers, if the coupling input to their corresponding subsystem exceeds a given bound. As a consequence, the communication topology is adjusted to the current control task and presents an image of the acting disturbances at all times. It is proven that the self-organizing controller leads to a practically stable overall system and that the effect of the disturbance can be bounded by an appropriate choice of the information activation condition. The approach is demonstrated by its application to a thermofluid process.