基于改进FNLM和HSEDO的振动筛轴承早期故障诊断  

Early fault diagnosis of vibrating screen bearings based on improved FNLM and HSEDO

作　　者：王艳[1] 叶敏[2] 李嘉波 张翠红 卜鹏辉 WANG YAN;YE Min;LI Jia-bo;ZHANG Cui-hong;BU Peng-hui(Xi'an Key Laboratory of Wellbore Integrity Evaluation,Xi'an Shiyou University,Xi'an 710065,Shaanxi,China;Key Laboratory of Expressway Construction Machinery of Shaanxi Province,Chang'an University,Xi'an 710061,Shaanxi,China)

机构地区：[1]西安石油大学西安市高难度复杂油气井完整性评价重点实验室,陕西西安710065 [2]长安大学道路施工技术与装备教育部重点实验室,陕西西安710061

出　　处：《长安大学学报(自然科学版)》2025年第1期167-178,共12页Journal of Chang’an University(Natural Science Edition)

基　　金：陕西省自然科学基础研究计划项目(2024JC-YBQN-0446);重点科研平台开放基金项目(300102253510);陕西省重点研发计划项目(2023-GHYB-05,2023-YBSF-104)。

摘　　要：针对振动筛滚动轴承早期故障特征微弱,被强背景噪声和多振源干扰淹没的诊断难题,提出一种基于改进快速非局部均值滤波(IFNLM)和高阶对称包络导数算子(HSEDO)的滚动轴承故障特征提取方法。首先,根据滚动轴承故障信号周期性结构特征,通过对快速非局部均值滤波(FNLM)算法相似度量和权值分配函数的改进,提出IFNLM算法,滤除轴承原始振动信号中的背景噪声和振源干扰;其次,在包络导数算子(EDO)的基础上,融入对称差分技术和高阶思想,提出HSEDO对降噪后的信号进行解调,从而识别轴承故障特征。最后,搭建振动筛激振器滚动轴承故障试验台,对所提方法的可行性和优越性进行验证。研究结果表明:IFNLM算法具有良好的去噪效果、噪声鲁棒性以及运行效率,IFNLM算法的运行速度相较于改进前提高了77%,约是小波阈值降噪算法的12倍,中值滤波算法的4倍;HSEDO算法具有良好的解调特性和抗干扰能力,在谐波干扰不高于20种的情况下,HSEDO算法处理的信号干扰比高于EDO;所提IFNLM-HSEDO算法与NLM+包络谱、WTD+TEO等现有方法相比,具有更高的故障特征提取能力和执行效率,在取50000个采样点运行时,耗时仅为0.0508 s,且采样点越多其快速性能越突出。研究成果对丰富机械系统故障诊断理论,保证机械设备安全可靠运行具有重要意义。A fault feature extraction method of rolling bearings based on improved fast nonlocal mean filtering(IFNLM)and high-order symmetric envelope derivative operator(HSEDO)was proposed to address the diagnostic challenge that weak early fault features of vibrating screen rolling bearings often being overwhelmed by heavy background noise and multiple vibration sources.Firstly,based on the periodic structural characteristics of the rolling bearing fault signals,an IFNLM algorithm was proposed by improving the similarity measure and weight assignment function of FNLM,which can filter the background noise and vibration source interference in the original vibration signals of the bearing.Secondly,based on the envelope derivative operator(EDO),integrating the symmetric difference technique and higher order idea,the HSEDO was proposed to demodulate the denoised signal and identify the bearing fault characteristics.Finally,the feasibility and superiority of the proposed method were verified by the vibrating screen exciter rolling bearing fault test rig.The research results show that the IFNLM algirithm has a good noise removal effect,noise robustness,and operation efficiency.The running speed of the IFNLM algorithm after improvement is 77%higher than before,about 12 times the wavelet threshold denoising algorithm and 4 times the median filtering algorithm.The HSEDO algorithm has outstanding demodulation and anti-interference capabilities,and its performance is superior to that of the EDO under no more than 20 types of harmonic interference.Compared with existing methods such as the NLM+envelope spectrum and WTD+TEO,the proposed IFNLM-HSEDO algorithm has higher fault feature extraction capability and execution efficiency.When 50000 sampling points are taken,the running time is only 0.0508 s,and the more sampling points,the more prominent the superiority of its fast performance.The research results are of great significance to enrich the theory of mechanical system fault diagnosis and ensure the safe and reliable operation of t

关 键 词：机械工程 振动筛轴承 故障诊断 IFNLM HSEDO 

分 类 号：U415.5[交通运输工程—道路与铁道工程]