Data-Driven Voltage Control in Isolated AC Microgrids Subject to Sensor Saturation

This paper proposes a distributed data-driven adaptive iterative learning control (DAILC) method to address the challenges of modeling and voltage regulation in isolated AC microgrids subject to disturbances and sensor saturation. Firstly, by utilizing the input–output data from the microgrid, a time-varying linear data microgrid model is developed for the distributed renewable energy generation unit (DREGU) that is independent of the microgrid’s physical information. Subsequently, the DAILC algorithm is developed from the microgrid data model, which only uses the input data and the corresponding saturated outputs from each DREGU, along with embedded measurement disturbances. Additionally, we verify the convergence of this algorithm, demonstrating that it can ensure the error in voltage restoration converges to a small neighborhood around the origin, even under conditions of sensor saturation and disturbances. Finally, through a simulation involving four DREGU nodes, we confirm the effectiveness of the proposed distributed DAILC method.