Scholarlink

The increasing prevalence of non-linear loads in residential and industrial microgrids undermines system stability by introducing current and voltage harmonics, which degrade power quality and increase thermal stress, leading to equipment malfunctions and reduced lifespan. Shunt active power filters are commonly employed for harmonic mitigation, but their effectiveness hinges on precise harmonic identification. This study presents a novel control scheme for four-wire microgrids that integrates generalized P-Q theory with an Adaptive Linear Neuron (ADALINE) filter. The proposed hybrid approach is evaluated in MATLAB Simulink under a heavily unbalanced scenario, involving both three-phase and single-phase non-linear loads, as well as distorted source voltage. Results show that the ADALINE filter significantly outperforms traditional passive filters in identifying and compensating harmonics, particularly under distorted conditions. Our method achieves superior harmonic mitigation, evidenced by a notable reduction in total harmonic distortion (THD), improved source current balancing, and enhanced system stability. This highlights the precision and robustness of the ADALINE-based approach compared to conventional low-pass filters in generalized instantaneous power theory.

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

Study of Robust Generalized PQ Theory Based on Artificial Neural Network for Unbalanced System

  • 会议论文

  • 发表日期：

    2026年

查看全文
  • Abdessalem Ryad Mebarek,
  • Leila Merabet,
  • Chouaib Rahli,
  • Saad Saleh

摘要

The increasing prevalence of non-linear loads in residential and industrial microgrids undermines system stability by introducing current and voltage harmonics, which degrade power quality and increase thermal stress, leading to equipment malfunctions and reduced lifespan. Shunt active power filters are commonly employed for harmonic mitigation, but their effectiveness hinges on precise harmonic identification. This study presents a novel control scheme for four-wire microgrids that integrates generalized P-Q theory with an Adaptive Linear Neuron (ADALINE) filter. The proposed hybrid approach is evaluated in MATLAB Simulink under a heavily unbalanced scenario, involving both three-phase and single-phase non-linear loads, as well as distorted source voltage. Results show that the ADALINE filter significantly outperforms traditional passive filters in identifying and compensating harmonics, particularly under distorted conditions. Our method achieves superior harmonic mitigation, evidenced by a notable reduction in total harmonic distortion (THD), improved source current balancing, and enhanced system stability. This highlights the precision and robustness of the ADALINE-based approach compared to conventional low-pass filters in generalized instantaneous power theory.