基于ITS与多重分形的传动箱状态识别

State Recognition of Vehicle Transmission Box Based on Both ITS and Multi-fractal

实验采集的传动箱振动信号受噪声干扰严重,难以直接通过信号的多重分形特征参量来区分不同工况,为此,采用ITS降噪及微弱特征提取方法,计算了降噪后传动箱正常运转、润滑不良及基座松动3种状态下的多重分形谱参数,对特征提取后信号从局部到整体、从渐进态到最终态的微变化过程进行了定量描述,并由这些特征值的敏感性判断了系统状态类别。

Because the vibration signals of transmission box collected from actual test is seriously interfered by noise,different working conditions cannot be distinguished through the multi-fractal characteristics directly.So ITS noise reducing and weak feature extraction method are used to calculate multi-fractal spectrum parameters of transmission box under normal operation,poor lubrication and pedestal looseness states.Quantitative description of the micro change process of signals ranging from partial to whole,from the progressive state to the final state are given after feature extraction,and the system status categories are judged by the eigenvalue sensitivity.