摘要
传统风电场风机变桨后备电源回路故障检测主要依赖人工巡检和定期检测,精度低且难以发现潜在故障。为此,提出快速检测方法,通过部署多层级故障检测节点,实现电源系统全面实时监控。引入高性能AD采样技术和局部检测法,精准捕获并分析故障数据,提取故障特征值。结合循环神经网络架构对故障进行分类识别,可显著提高检测精度和速度。实验证明,该方法在风电场风机变桨后备电源回路故障排查中展现出显著优势。
Traditional wind farm wind turbine pitch backup power circuit fault detection mainly relies on manual inspection and regular testing,with low accuracy and difficulty in detecting potential faults.Therefore,a fast detection method is proposed to achieve comprehensive real-time monitoring of the power system by deploying multi-level fault detection nodes.High-performance AD sampling technology and local detection methods are introduced to accurately capture and analyze fault data,and extract fault feature values.Combining recurrent neural network architecture for fault classification and recognition can significantly improve detection accuracy and speed.The experimental results have shown that this method exhibits significant advantages in troubleshooting the backup power circuit of wind turbines pitch in wind farms.
作者
东学军
徐振北
赵成利
DONG Xuejun;XU Zhenbei;ZHAO Chengli(Xintian Green Energy Weichang Co.,Ltd.,Chengde,Hebei 068450,China)
出处
《自动化应用》
2025年第15期125-127,共3页
Automation Application
关键词
风电场
风机变桨
后备电源回路
回路故障
故障检测
快速检测
wind farm
wind turbine pitch
backup power circuit
circuit faults
fault detection
rapid detection
