Black-box Identification for an auto-tuned current controller working with Voltage Source Converters connected to the grid through a LCL filter

This work proposes a novel method based on a prediction error minimization (PEM) algorithm to identify the discrete state-space model of a LCL filter which is used as interface between a Voltage Source Converter and the utility grid. This model is used to design directly a discrete state-space current control based on a servo controller. The so-obtained model makes easier the use of advanced control techniques as Linear Quadratic Regulators, state estimators, Robust Control, etc. The main advantage of using an identification process against to design the controller starting from the manufacturer data is that the control is not sensitive to modeled errors. The method can be applied to closed-loop systems, being suitable for adaptive control designs.