基于叶片加速度振动信号的风机叶片状态监测系统  

A condition monitoring system for fan blades based on blade acceleration vibration signal

作　　者：蔺雪峰 赵江 孟秀俊 苗雷 李缓 Lin Xuefeng;Zhao Jiang;Meng Xiujun;Miao Lei;Li Yuan(China Huaneng Group Clean Energy Technology Research Institute Co.,Ltd.,Beijing,100031,China;Huaneng Guizhou Clean Energy Branch Co.,Guizhou Guiyang,550081,China;Shenyang Shengshi Renewable Energy Technology Co.,Ltd.,Liaoning Shenyang,110079,China)

机构地区：[1]中国华能集团清洁能源技术研究院有限公司,北京100031 [2]华能国际电力股份有限公司贵州清洁能源分公司,贵州贵阳550081 [3]沈阳盛世新能源科技有限公司,辽宁沈阳110079

出　　处：《机械设计与制造工程》2024年第5期96-100,共5页Machine Design and Manufacturing Engineering

基　　金：2022年度中国华能集团总部科技项目(HNKJ22-H101)。

摘　　要：针对叶片振动信号噪声导致叶片状态监测结果存在误差的问题,基于叶片振动加速度分量分析叶片振动信号,设计一种风机叶片状态实时监测系统。设计数据采集方案以及实验参数,对采集到的风机叶片状态数据通过短时傅里叶变换和小波包进行预处理和特征提取,运用三层小波包分解方式分析提取特征,通过支持向量机进行故障分类与识别,获取故障诊断结果。实验测试结果表明,所设计系统监测的叶片位移结果与实际结果较为吻合,能够根据叶片振动信号实现对叶片状态的监测和故障诊断。A real-time monitoring system for fan blade status is designed based on the analysis of blade vibration acceleration components to address the issue of errors in blade status monitoring caused by noise in blade vibration signals.It designs a data collection plan and experimental parameters,preprocess and extract features from the collected wind turbine blade state data through short-time Fourier transform and wavelet packet,analyzes the extracted features through three-layer decomposition of wavelet packet,and uses support vector machine for fault classification and recognition to obtain fault diagnosis results.The experimental test results show that the displacement results monitored by the designed system are in good agreement with the actual results,and can achieve monitoring of blade status and fault diagnosis based on blade vibration signals.

关 键 词：风机 风机叶片 加速度传感器 振动信号处理 故障诊断 

分 类 号：TP23[自动化与计算机技术—检测技术与自动化装置]