Model Predictive Control (MPC) is widely used in control systems for its ability to handle constraints and optimize performance. However, conventional MPC methods often employ Euler integration for trajectory computation, which introduces computational errors that escalate with the sampling time, leading to diminished tracking performance and higher switching frequencies, particularly at larger intervals. To address these challenges, we propose a novel approach that integrates the 4th-order Runge-Kutta (4oRK) method into MPC. The 4oRK method offers improved accuracy over Euler integration by significantly reducing computational errors through its higher-order approximation. A comparative analysis of the two methods, conducted under varying load profiles and voltage sag conditions, revealed that while the Euler-based approach produces grid currents with a Total Harmonic Distortion (THD) exceeding 5 %, the 4oRK-based method consistently achieves a THD below 5 %, ensuring superior harmonic suppression. Moreover, the 4oRK method effectively reduces power losses without increasing computational complexity, as demonstrated by comparable task execution times. This improvement is achieved through a two-stage computation process prediction and correction that enhances MPC's performance at larger sampling intervals while reducing control adjustment frequency. Extensive MATLAB/Simulink simulations and physical prototype experiments validate the proposed 4oRK-based MPC, showing its ability to minimize THD, achieve unity power factor, and maintain robust control performance at lower switching frequencies. This advancement in MPC integration contributes to more efficient, accurate, and reliable control system design.
扫码关注我们
求助内容：
应助结果提醒方式：
京公网安备 11010802042870号