排列熵与核极限学习机在齿轮故障诊断中的应用

发布时间：2019-01-09 16:25

【摘要】：针对齿轮故障难提取和极限学习机(extreme learning machine,ELM)隐层节点数需要人为设定,致使齿轮故障分类模型准确度低、稳定性差的问题,提出基于核极限学习机(kernel extreme learning machine,K-ELM)的齿轮故障诊断方法。首先,将测得信号经经验模态分解(empirical mode decomposition,EMD)处理后得到一系列IMF本征模式分量,并提取各分量的排列熵(permutation entropy,PE)值组成高维特征向量集;然后利用高斯核函数的内积表达ELM输出函数,从而自适应确定隐层节点数;最后,将所得高维特征向量集作为K-ELM算法的输入建立核函数极限学习机齿轮故障分类模型,进行齿轮不同故障状态的分类辨识。实验结果表明:与SVM、ELM故障分类模型相比,核函数ELM滚动齿轮故障诊断分类模型具有更高的准确度和稳定性。
[Abstract]:Aiming at the problem that the hidden node points of gear fault extraction and (extreme learning machine,ELM) need to be set artificially, the accuracy of gear fault classification model is low and the stability of gear fault classification model is poor, a new method based on kernel limit learning machine (kernel extreme learning machine,) is proposed. The method of gear fault diagnosis based on K-ELM. First, a series of IMF eigenmode components are obtained after the measured signal is processed by empirical mode decomposition (empirical mode decomposition,EMD), and the permutation entropy (permutation entropy,PE) values of each component are extracted to form a high dimensional eigenvector set. Then the inner product of Gao Si kernel function is used to express the ELM output function, which adaptively determines the number of hidden layer nodes. Finally, the high dimensional eigenvector set is used as the input of K-ELM algorithm to establish the kernel function extreme learning machine gear fault classification model, and to classify and identify the different fault states of gear. The experimental results show that the kernel function ELM rolling gear fault diagnosis model has higher accuracy and stability than the SVM,ELM fault classification model.
【作者单位】：内蒙古科技大学机械工程学院;
【基金】：国家自然科学基金(51565046) 内蒙古自然科学基金(2015MS0512) 内蒙古科技大学创新基金(2015QDL12)
【分类号】：TH132.41

【参考文献】