Improved Model Predictive Control Based on Luenberger State Observer for LC-Coupling Hybrid Active Power Filter

An improved model predictive control (MPC) with Luenberger state observer is proposed in this article for LC-coupling hybrid active power filter (LC-HAPF). Generally, the implementation of MPC for LC-HAPF requires additional sensors for measuring coupling capacitor voltage, which increases the system cost and reduces reliability. To relax these problems, both continuous and discrete models are deduced first. Then, the Luenberger observer is utilized based on the discrete model to mitigate the steady-state error without additional sensors. In addition, the delay compensation is given for the digital control delay. Finally, the effectiveness of the improved MPC for LC-HAPF is verified by experiment results, indicating that the proposed method has fast dynamic response and low steady-state error.