Controller identification for data-driven model-reference distributed control

This paper considers data-driven distributed controller synthesis for interconnected linear systems subject to unmeasured disturbances. The considered problem is the op-timization of a model-reference control criterion, where the reference model is described by a decoupled system. We provide a method to determine the optimal distributed controller by performing network identification in an augmented network. Sufficient conditions are provided for which the data-driven method solves the distributed model-reference control problem, whereas state-of-the-art methods for data-driven distributed control can only provide performance guarantees in the absence of disturbances. The effectiveness of the method is demonstrated via a simple network example consisting of two interconnected systems.