Abstract:

While in the background of strong noise, it is difficult to extract the non-stationary nonlinear signal feature of rolling bearing. Therefore, this paper proposes a method based on Local Characteristic-scale Decomposition (LCD) and Kernel Extreme Learning Machine (KELM) intelligent diagnostic method (LCD_KELM). Firstly, the measured vibration signals are processed with LCD and decomposes into a series of intrinsic scale component (ISC). Then a number of ISCs containing valid information components are selected out and then compute their ApEns. Using KELM to train randomly selected ApEns. Using the remaining ApEns to test after saving the training parameters and the experimental results show that LCD_KELM has a high diagnostic accuracy, it can accurately diagnosis of rolling bearing operating status to determine the running state of the rolling bearing.

摘要：

由于提取滚动轴承的非平稳非线性信号特征较为困难，强噪声背景下难以诊断早期故障，故而提出一种基于局部特征尺度分解(Local Characteristic-scale Decomposition，LCD)和核极限学习机(Kernel Extreme Learning Machine，KELM)的智能诊断方法（LCD_KELM）。该方法首先对信号进行LCD分解，将其分解成多个內禀尺度函数（Intrinsic Scale Component ,ISC），选取包含有效频率成分的ISC并计算其近似熵值(Approximate Entropy，ApEn)，使用KELM对随机选取的近似熵值进行训练，保存训练参数后，利用剩余的近似熵值进行测试，实验结果表明LCD_KELM具有较高的诊断准确率，能够对滚动轴承运行状态进行高精度诊断，从而判断滚动轴承的运转状况。

关键词: 振动与波, 故障诊断, 故障信号提取, 局部特征尺度分解, 近似熵, 核极限学习机