Subspace identification of distributed, decomposable systems

This article concerns the identification of a class of linear systems which we call ¿decomposable systems¿. Such systems can be thought of as the interconnection of a number of identical subsystems, and they can be used to model a number of large scale systems. We propose an algorithm that allows the identification of such systems keeping their structure in the resulting model. The algorithm is made by coupling a subspace identification method with a bilinear matrix inequality (BMI) that can be solved iteratively.