基于人工智能的风机叶片故障自动化监测系统  

Automated monitoring system for fan blade faults based on artificial intelligence

作　　者：闫浩伟 YAN Haowei(Datang Shanxi Renewable Power Company,Taiyuan 030000,China)

机构地区：[1]大唐山西新能源公司,山西太原030000

出　　处：《电子设计工程》2025年第8期87-91,共5页Electronic Design Engineering

摘　　要：针对风机叶片长期暴露于恶劣环境且承受巨大应力,其健康状况监测不足导致故障发现滞后,影响风机性能和发电量的问题,设计基于人工智能的风机叶片故障自动化监测系统。通过集成多种设备的叶片数据采集模块实时监测叶片运行状态,利用模态动能法优化布置模块,采用卷积神经网络与短时傅里叶变换相结合的故障诊断模型进行诊断。最后以4.7m冷却塔风机为实验对象,进行了风机叶片故障诊断仿真实验,结果表明,该系统能实现叶片结冰故障诊断与定位,且对结冰、裂纹故障及正常信号分类性能优于对比方法。设计系统能够高精度实现叶片故障诊断,并获取风机叶片故障点位,具有较高的实际应用价值。Aiming at the problem that the fan blades are exposed to harsh environment for a long time and bear great stress,and the lack of health monitoring leads to the lag of fault detection,which affects the performance and power generation of the fan,an automatic monitoring system for fan blade faults based on artificial intelligence is designed.The operation state of blades is monitored in real time by integrating blade data acquisition modules of various devices,and the layout of modules is optimized by modal kinetic energy method.The fault diagnosis model combining convolutional neural network and short time Fourier transform is used for diagnosis.Finally,the simulation experiment of fan blade fault diagnosis is carried out,with a 4.7 m cooling tower fan as the experimental object.The test results show that the system can realize the diagnosis and location of blade icing fault,and the classification performance of icing,crack fault and normal signal is better than the comparison method.The design system can realize blade fault diagnosis with high precision and obtain the fault point of fan blade,which has high practical application value.

关 键 词：太阳能供电电路 模态动能法 风机叶片故障监测 卷积神经网络 短时傅里叶变换 

分 类 号：TN919[电子电信—通信与信息系统]