基于优化EFD算法的风电行星齿轮箱故障诊断研究  

Wind Power Planetary Gearbox Fault Diagnosis Based on Optimized EFD Algorithm

作　　者：王国锋[1] 张旭东 汪菲[1] 户满堂 Wang Guofeng;Zhang Xudong;Wang Fei;Hu Mantang(School of Mechanical Engineering,Tianjin University,Tianjin 300350,China)

机构地区：[1]天津大学机械工程学院,天津300350

出　　处：《天津大学学报（自然科学与工程技术版）》2023年第4期355-360,共6页Journal of Tianjin University：Science and Technology

基　　金：国家重点研发计划资助项目(2019YFB1704802-2,2019YFA0706702);国家自然科学基金资助项目(52075365).

摘　　要：行星齿轮箱作为风力发电机的关键核心部件,对其故障进行准确诊断能有效提升风力发电效能.行星轮系作为一种复杂的传动机械部件,其频谱表现异常复杂,且故障信息极易被无关成分或干扰成分淹没,而利用信号分解获取故障分量的方法在行星齿轮故障诊断中发挥着重要的作用.因此,针对经验傅里叶分解(empirical Fourier decomposition,EFD)易陷入局部频谱分割的问题,优化改进了EFD的频谱分割算法,即在原频谱分割算法上引入边界阈值机制,优化频谱分割边界点的选择,有效限制边界频率陷入局部的问题.通过构造多分量仿真信号对比分析原频谱分割算法和优化算法,并逐步增加分量成分对比分析.仿真分析结果表明,原频谱分割算法随着分量成分的增加,其边界频率逐渐陷入局部,而优化算法却能准确获取边界频率,验证了优化EFD算法的有效性,表明优化频谱分割算法是在原频谱分割算法上的有效改进.最后通过对风电行星齿轮箱实验数据的分析表明,与EFD算法相比,优化EFD算法获取的边界频率不易陷入局部,可以更好地获取故障分量.在对风电行星齿轮箱的故障诊断中,能更有效地识别故障频率成分和确定故障位置.As the key core component of wind turbines,planetary gearboxes can effectively improve wind power generation efficiency by diagnosing their faults accurately.Being an intricate mechanical component of transmission,the planetary gear system exhibits an exceptionally complex spectral performance,the fault information is easily overwhelmed by irrelevant or disturbing components,and using signal decomposition to obtain fault components plays an important role in planetary gearbox fault diagnosis.Therefore,to address the problem of empirical Fourier decomposition(EFD)easily falling into the local spectrum segmentation,its spectrum segmentation algorithm is op-timized,and the boundary threshold mechanism is introduced to the original spectrum segmentation algorithm to optimize the selection of the boundary point of spectrum segmentation and effectively limit the boundary frequency falling into the local problem.A multi-component simulation signal is constructed to compare and analyze the original and the optimized spectrum segmentation algorithms,and the components for the comparison and analysis are gradu-ally increased.The simulation analysis results show that the original spectrum segmentation algorithm gradually falls into the local boundary frequency with the increase of components,while the optimized algorithm can accurately obtain the boundary frequency,thus verifying the effectiveness of the optimized EFD algorithm and showing that the optimized algorithm is an effective improvement on the original spectrum segmentation algorithm.Finally,the experimental data analysis of the wind power planetary gearbox shows that compared with the EFD algorithm,the boundary frequency obtained by the optimized EFD algorithm is not easy to fall into the local area,and the fault components can be better obtained.In the fault diagnosis of the wind power planetary gearbox,it can more effectively identify the fault frequency components and determine the fault location.

关 键 词：经验傅里叶分解 故障诊断 频谱分割 行星齿轮箱 

分 类 号：TH17[机械工程—机械制造及自动化]