Scholarlink

<p>Maintaining high power quality in wind energy conversion systems (WECS) is challenging due to voltage fluctuations and harmonics. This research proposes a novel COPO-LRNN approach, integrating chaotic opposition-based parrot optimizer (COPO) with a local randomized neural network (LRNN) to control a quasi-Z-source inverter (QZSI) using a fractional-order tilting integral double derivative (FOTIDD<sup>2</sup>) controller. COPO optimizes controller gains, while LRNN predicts precise inverter signals, enhancing voltage regulation, efficiency, and system reliability. MATLAB results show a voltage THD of 0.6%, efficiency of 99%, low voltage sag (15.3%), minimal voltage swell (4.23%), and a settling time of 0.2&#xa0;s, outperforming existing optimization methods like gorilla troops algorithm (GTA), genetic-based chicken swarm algorithm (GBCSA), and particle swarm optimization (PSO). The proposed COPO-LRNN strategy provides a robust, efficient, and high-performance solution for superior power quality in WECS.</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

Control of PMSG-based wind turbines with quasi-Z-source inverter and fractional-order tilting integral double-derivative controller for performance enhancement under fluctuating wind and grid conditions

  • 期刊论文

  • 发表日期：

    2026年5月27日

查看全文
  • E. Vani,
  • P. Balakrishnan,
  • Jayakumar Thangavel,
  • S. Senthilkumar

摘要

Maintaining high power quality in wind energy conversion systems (WECS) is challenging due to voltage fluctuations and harmonics. This research proposes a novel COPO-LRNN approach, integrating chaotic opposition-based parrot optimizer (COPO) with a local randomized neural network (LRNN) to control a quasi-Z-source inverter (QZSI) using a fractional-order tilting integral double derivative (FOTIDD2) controller. COPO optimizes controller gains, while LRNN predicts precise inverter signals, enhancing voltage regulation, efficiency, and system reliability. MATLAB results show a voltage THD of 0.6%, efficiency of 99%, low voltage sag (15.3%), minimal voltage swell (4.23%), and a settling time of 0.2 s, outperforming existing optimization methods like gorilla troops algorithm (GTA), genetic-based chicken swarm algorithm (GBCSA), and particle swarm optimization (PSO). The proposed COPO-LRNN strategy provides a robust, efficient, and high-performance solution for superior power quality in WECS.