Scholarlink

<p>Global wind energy production is growing yearly, representing a shift towards renewable energy sources with a low carbon footprint. In this context, Industry 4.0 technologies can provide opportunities for small-scale wind power generation units to enhance production and machine efficiency. Small-scale wind turbines typically lack fault detection systems for issues such as mass imbalance, which can result from cracked blades or icing. These faults may take a long time to detect and can significantly reduce energy production efficiency. In this scenario, Intelligent Fault Diagnosis (IFD) systems can improve the reliability and safety of generating units, preventing production losses, increasing equipment availability, and reducing corrective maintenance costs. Thus, this work presents a method for detecting imbalance faults in wind turbines based on easy-to-measure signals, such as electric current and voltage. Two different artificial intelligence strategies are investigated: the Machine Learning approach based on artificial neural networks and the Deep Learning approach based on convolutional neural networks. The results show that the methodology is feasible, reaching high accuracy rates: 98.0% for the machine learning approach and 100.0% for the deep learning approach. Therefore, this methodology is reliable and non-intrusive and represents a cost-effective solution for micro- and mini-wind power generation systems.</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

Intelligent fault diagnosis of rotor imbalance for small-scale wind turbines based on easy-to-measure signals

  • 期刊论文

  • 出版模式：

    开放获取

  • 发表日期：

    2026年3月28日

查看全文
  • Andre Luis Dias,
  • Willian Kenji Ishioka,
  • Guilherme Serpa Sestito,
  • Rodrigo Nicoletti

摘要

Global wind energy production is growing yearly, representing a shift towards renewable energy sources with a low carbon footprint. In this context, Industry 4.0 technologies can provide opportunities for small-scale wind power generation units to enhance production and machine efficiency. Small-scale wind turbines typically lack fault detection systems for issues such as mass imbalance, which can result from cracked blades or icing. These faults may take a long time to detect and can significantly reduce energy production efficiency. In this scenario, Intelligent Fault Diagnosis (IFD) systems can improve the reliability and safety of generating units, preventing production losses, increasing equipment availability, and reducing corrective maintenance costs. Thus, this work presents a method for detecting imbalance faults in wind turbines based on easy-to-measure signals, such as electric current and voltage. Two different artificial intelligence strategies are investigated: the Machine Learning approach based on artificial neural networks and the Deep Learning approach based on convolutional neural networks. The results show that the methodology is feasible, reaching high accuracy rates: 98.0% for the machine learning approach and 100.0% for the deep learning approach. Therefore, this methodology is reliable and non-intrusive and represents a cost-effective solution for micro- and mini-wind power generation systems.