A Distributed Compensation Method for Active Power Filters Based on Model Predictive Control and Kalman Filter Algorithm

This project looks at a distributed compensation method for active power filters (APFs) that reduces overall voltage distortion in modern distribution networks with many distributed nonlinear loads. Instead of fixing the problem locally, the approach coordinates a small group of APFs whose reference signals can change dynamically, aiming to improve power quality across the whole network. The control system combines Model Predictive Control (MPC) with a Kalman Filter (KF) state estimator. The MPC part predicts how the system will behave over a short future horizon and calculates the best current references for each filter, targeting lower voltage distortion at important buses. At the same time, the Kalman Filter estimates the

initial harmonic voltages, removing measurement noise and handling uncertainties in the system model. This two‑loop strategy keeps performance stable even when loads vary. The method is tested on a multi‑bus distribution network model—such as a seven‑bus system—using MATLAB/Simulink simulations. The evaluation checks three things: how much total harmonic distortion (THD) is reduced, how accurately the predictive controller follows its targets, and how well the Kalman estimator rejects noise. Together, these results provide a solid framework for advanced power‑quality management in smart grids.

Message: Hi FE, I need help with “A Distributed Compensation Method for Active Power Filters Based on Model Predictive Control and Kalman Filter Algorithm” in “Electrical Engineering”