IHIE在滚动轴承损伤识别中的应用  

Application of IHIE in damage identification of rolling bearing

作　　者：孟秋静 杨钢[2] 王红卫 MENG Qiujing;YANG Gang;WANG Hongwei(Sino-German Institute of Engineering,Shanghai Technical Institute of Electronics&Information,Shanghai 201411,China;School of Mechanotronics and Vehicle Engineering,Chongqing Jiaotong University,Chongqing 400074,China;School of Mechanical Engineering,Southeast University,Nanjing 210096,Jiangsu,China)

机构地区：[1]上海电子信息职业技术学院中德工程学院,上海201411 [2]重庆交通大学机电与车辆工程学院,重庆400074 [3]东南大学机械工程学院,江苏南京210096

出　　处：《中国工程机械学报》2023年第5期476-481,497,共7页Chinese Journal of Construction Machinery

基　　金：国家自然科学基金资助项目(52075095);安徽省高校重点自然科学基金资助项目(KJ2017A781)。

摘　　要：针对滚动轴承的损伤识别精度较低的缺陷,建立基于改进层次增量熵(IHIE)和海鸥算法(SOA)优化极限学习机(ELM)的损伤识别模型。首先,采用IHIE提取滚动轴承的熵值特征,生成故障特征;然后,利用SOA对ELM进行迭代寻优,建立结构最优的分类模型;最后,将故障特征输入至SOA-ELM模型中进行训练和测试,完成滚动轴承的损伤识别。利用两组滚动轴承数据集对该模型进行了实验评估。研究结果表明:该模型能够有效识别滚动轴承的损伤类型和损伤程度,两种数据集的识别准确率分别达到了100%和96.92%,且在多个维度都优于对比方法,具有一定的优越性。Aiming at the defects of low damage identification accuracy of rolling bearings,a damage identification model based on improved hierarchical increment entropy(IHIE)and seagull optimization algorithm(SOA)optimized extreme learning machine(ELM)was established.Firstly,IHIE was used to extract the entropy characteristics of rolling bearings and generate fault feature.Then,SOA was used to iteratively optimize ELM,and an optimal classification model was established.Finally,the fault characteristics were input into the SOA-ELM model for training and testing to complete the damage identification of rolling bearings.Two sets of rolling bearing data sets were used to evaluate the model experimentally.The research results show that the model can effectively identify the damage type and damage degree of rolling bearing,and the recognition accuracy of the two data sets reaches 100%and 96.92%respectively,and is superior to the comparison method in many dimensions,with certain advantages.

关 键 词：改进层次增量熵(IHIE) 极限学习机(ELM) 损伤识别 滚动轴承 

分 类 号：TH133.3[机械工程—机械制造及自动化]