Scholarlink

<p>A Wind Energy Conversion System (WECS) operates with a Permanent Magnet Synchronous Generator (PMSG) and a Single-Ended Primary Inductor Converter (SEPIC), which delivers effective energy conversion because it enables better control of generated energy. The system experiences performance degradation because of generator and converter losses that occur, especially when wind speeds decrease. The output current experiences increased Total Harmonic Distortion (THD) because the SEPIC converter switching dynamics affect its performance. This paper introduces a hybrid solution that combines Light Spectrum Optimizer (LSO) with Pyramidal Dilation Attention Convolutional Neural Network (PDACNN) to solve these problems in WECS. The proposed method targets two main objectives: decreasing THD values and increasing system operational efficiency. The proposed LSO optimizes operational parameters through duty cycle and switching frequency (SF) control to achieve effective energy conversion in the wind energy system. The PDACNN system uses its wind speed and generator output predictions to enable precise control adjustments that maintain system stability while achieving optimal performance results. The research used existing methods such as Fuzzy Logic Control (FLC), Non-Dominated Sorting Genetic Algorithm III-Finite Element Analysis (NSGA III-FEA), Jumping Spider Optimization Algorithm-Pelican Optimization Algorithm-Teaching Learning Studying based Optimization (JSOA-POA-TLSBO), Improved Honey Badger Algorithm (IHBA), and Back Propagation Artificial Neural Network-Genetic Algorithm-Particle Swarm Optimization (BPANN-GA-PSO) to compare the proposed method implemented in MATLAB. The developed model demonstrates better performance compared to current methodologies because it achieves a total harmonic distortion of 0.82% and a system efficiency of 98.8%. The technique demonstrates its effectiveness for wind energy conversion because it achieves three goals: reducing harmonic distortion, increasing efficiency, and maintaining dependable wind energy conversion. </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

Light spectrum optimizer and pyramidal dilation attention convolutional neural network for wind energy conversion system using permanent magnet synchronous generator and single-ended primary inductor converter

  • 期刊论文

  • 出版模式：

    开放获取

  • 发表日期：

    2026年6月25日

查看全文
  • K. R. Suja,
  • M. SivaramKrishnan,
  • J. Isaac JoshuaRamesh Lalvani,
  • S. SriRagavi

摘要

A Wind Energy Conversion System (WECS) operates with a Permanent Magnet Synchronous Generator (PMSG) and a Single-Ended Primary Inductor Converter (SEPIC), which delivers effective energy conversion because it enables better control of generated energy. The system experiences performance degradation because of generator and converter losses that occur, especially when wind speeds decrease. The output current experiences increased Total Harmonic Distortion (THD) because the SEPIC converter switching dynamics affect its performance. This paper introduces a hybrid solution that combines Light Spectrum Optimizer (LSO) with Pyramidal Dilation Attention Convolutional Neural Network (PDACNN) to solve these problems in WECS. The proposed method targets two main objectives: decreasing THD values and increasing system operational efficiency. The proposed LSO optimizes operational parameters through duty cycle and switching frequency (SF) control to achieve effective energy conversion in the wind energy system. The PDACNN system uses its wind speed and generator output predictions to enable precise control adjustments that maintain system stability while achieving optimal performance results. The research used existing methods such as Fuzzy Logic Control (FLC), Non-Dominated Sorting Genetic Algorithm III-Finite Element Analysis (NSGA III-FEA), Jumping Spider Optimization Algorithm-Pelican Optimization Algorithm-Teaching Learning Studying based Optimization (JSOA-POA-TLSBO), Improved Honey Badger Algorithm (IHBA), and Back Propagation Artificial Neural Network-Genetic Algorithm-Particle Swarm Optimization (BPANN-GA-PSO) to compare the proposed method implemented in MATLAB. The developed model demonstrates better performance compared to current methodologies because it achieves a total harmonic distortion of 0.82% and a system efficiency of 98.8%. The technique demonstrates its effectiveness for wind energy conversion because it achieves three goals: reducing harmonic distortion, increasing efficiency, and maintaining dependable wind energy conversion.