Scholarlink

<p>In practice, shunt active power filters (SAPFs) are widely employed with nonlinear loads to improve power quality, including the minimisation of harmonics and reactive power compensation. When the source voltage contains distortion, unbalanced magnitude, or other irregularities, conventional SAPF control systems have utilised a high-selectivity filter (HSF) or a dual second-order generalised integrator (DSOGI), which are limited in adaptability and estimation accuracy. To improve its accuracy, the proposed control architecture integrates the unscented Kalman filter (UKF), replacing HSF or DSOGI. The UKF provides a more accurate yet faster estimate of the fundamental component of the source voltage, which is essential for subsequent control actions. Furthermore, here the model predictive direct power control (MPDPC) is used for switching control, and adaptive back-stepping control (BSC) for adaptive reactive power reference within a unified SAPF framework. The MPDPC leverages its predictive capabilities to achieve improved control performance, replacing static switching approaches. Moreover, the back-stepping-based adaptive reactive power generation scheme enhances control accuracy during grid disturbances. Simulation and real-time software-in-the-loop validation using an OP-5142 digital simulator confirms that the proposed system achieves a reduction in source-current THD from 3.36 to 2.97% under balanced and from 4.05 to 3.27% under unbalanced grids, a 20% decrease in reactive-power ripple, and an 83% improvement in steady-state reactive-power accuracy, with a settling time below 10&#xa0;ms and near-unity power factor (≈ 0.998). These results demonstrate the proposed UKF–MPDPC–BSC-based SAPF’s superior harmonic suppression and robust real-time performance compared with conventional SAPF controllers.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Model predictive direct power control for shunt active power filter using adaptive back stepping control and unscented Kalman filter

  • 期刊论文

  • 发表日期：

    2026年4月18日

查看全文
  • Satyanarayan Padhy,
  • Kanhu Charan Bhuyan,
  • Abinash Rath

摘要

In practice, shunt active power filters (SAPFs) are widely employed with nonlinear loads to improve power quality, including the minimisation of harmonics and reactive power compensation. When the source voltage contains distortion, unbalanced magnitude, or other irregularities, conventional SAPF control systems have utilised a high-selectivity filter (HSF) or a dual second-order generalised integrator (DSOGI), which are limited in adaptability and estimation accuracy. To improve its accuracy, the proposed control architecture integrates the unscented Kalman filter (UKF), replacing HSF or DSOGI. The UKF provides a more accurate yet faster estimate of the fundamental component of the source voltage, which is essential for subsequent control actions. Furthermore, here the model predictive direct power control (MPDPC) is used for switching control, and adaptive back-stepping control (BSC) for adaptive reactive power reference within a unified SAPF framework. The MPDPC leverages its predictive capabilities to achieve improved control performance, replacing static switching approaches. Moreover, the back-stepping-based adaptive reactive power generation scheme enhances control accuracy during grid disturbances. Simulation and real-time software-in-the-loop validation using an OP-5142 digital simulator confirms that the proposed system achieves a reduction in source-current THD from 3.36 to 2.97% under balanced and from 4.05 to 3.27% under unbalanced grids, a 20% decrease in reactive-power ripple, and an 83% improvement in steady-state reactive-power accuracy, with a settling time below 10 ms and near-unity power factor (≈ 0.998). These results demonstrate the proposed UKF–MPDPC–BSC-based SAPF’s superior harmonic suppression and robust real-time performance compared with conventional SAPF controllers.