摘要
针对2D80–53.4型压缩机故障信号呈现高耦合性和非线性等问题,提出基于RCMDE的2D80–53.4型压缩机故障诊断方法。在MDE算法的基础上,结合精细复合改进MDE算法得到RCMDE算法,应用其对非线性振动信号进行特征向量构建,结合极限学习机进行故障识别。通过仿真信号验证,结果表明该算法可以有效抑制干扰信息,强化故障信息特征,大大提高了算法的准确性。以2D80–53.4型压缩机的轴承故障数据为研究对象进行实测验证,应用RCMDE实现其故障信号特征提取,与多尺度散布熵、复合多尺度散布熵进行对比,该方法表现出特征可分性良好,极限学习机故障识别准确率较高。
Aiming at the problems of high coupling and nonlinearity in bearing fault signals of 2D80–53.4 model machine,a method for 2D80–53.4 reciprocating compressor based on RCMDE algorithm is proposed.The RCMDE algorithm is obtained by improving MDE algorithm,and applied to construct the feature vector of the nonlinear vibration signal,then the fault identification is studied by an extreme learning machine.Simulation signal verification shows that the algorithm can effectively suppress interference information and strengthen the characteristics of fault information,thus the accuracy of the algorithm is greatly improved.The RCMDE is used to realize the feature extraction of the fault signal for the bearing fault data of the 2D80–53.4 reciprocating compressor,and compared with the multi-scale walking entropy and the compound multi-scale walking entropy.The results prove that the feature separability of this method is better,and the fault recognition accuracy of the extreme learning machine is higher.
作者
曲孝海
胡予欢
沈磊
Qu Xiaohai;Hu Yuhuan;Shen Lei(College of Mathematics and Physics,Hunan University of Arts and Science,Changde 415000,China)
出处
《湖南文理学院学报(自然科学版)》
CAS
2022年第2期8-12,共5页
Journal of Hunan University of Arts and Science(Science and Technology)
基金
湖南省教育厅科学研究项目(21C0518)。
关键词
2D80–53.4型压缩机
精细复合多尺度散布熵
极限学习机
故障诊断
2D80–53.4 reciprocating compressor
fine composite multiscale scattering entropy
extreme learning machine
fault diagnosis