Model Identification of Distributed Energy Resources Using Sparse Regression and Koopman Theory

A complex physics-based modeling procedure and the uncertainty and confidentiality of internal parameters of distributed energy resources (DERs) motivate system identification tools. With the availability of high-fidelity measurements and historical data, model-free identification of DERs can facilitate the control design without tedious modeling of these nonlinear systems. This paper develops a framework for data-driven nonlinear modeling of DERs using sparse identification of nonlinear dynamics (SINDy). In addition, to avoid the complexities of nonlinear control designs, the identified nonlinear dynamics will be lifted to a linear space utilizing Koopman theory. Compared with existing physics-based designs that heavily rely on knowing the detailed system dynamics or data-driven designs that relay on large historical data and are not interpretable, the proposed model-free DER identification framework can accurately capture the dynamics of the DERs with available measurements and lift them to a linear space that provides guaranteed performance for tracking problems. Time-domain simulations were carried out to validate the effectiveness of the proposed approach.