基于SOM网的风电变流器故障诊断

摘要

我国新疆、甘肃、宁夏、内蒙、浙江、黑龙江、江苏、广东等都在大规模建设风电场,这些风电场建成后,其故障维护就有了很大市场.以新疆风电场为基础,尝试开发用于风力机故障智能诊断的系统.首先介绍了风力机及其变频器系统的结构,分析了变频器的故障机理.使用SOM神经网络对风机变流器进行了诊断,用数据验证了诊断结果.把传统的电力电子设备故障诊断技术与新疆风力机变频器的故障诊断相结合,为风电大面积推广应用产生了积极作用.%Large-scaledwindpowerfarms are underconstruction at present in Xinjiang Gansu Ningxia Inner Magnolia,Heilongjiag Jiangsu Guangdong and other provinces and autonomous repons in China The completion and operation of these wind farms will create a huge market of wind farm maintenance service. This paper introduced a fault diagnosis system for converters of wind turbines based on wind farms in Xinjiang This paper firstly introduced the structure of the wind turbine and its converter and analyzed causes of converters' faults The SOM neural net was used to diagnose the faults and the data generated verified its effect The study innovatively combines the traditional fault diagnosis technology for electncal and electronic devices with the fault diagnosis technology for wind turbines'converters in Xinpang wind farms, and will play a positive role in popularization of the technique in more wind farms.