基于SPS与CNN的行星齿轮箱故障特征提取与诊断研究

Research of Fault Feature Extraction and Diagnosis of Planetary Gear Train based on SPS and CNN

针对行星齿轮箱结构和运行工况复杂,导致信号故障特征提取困难的问题,通过分析行星轮系振动机理,初步推导出含故障齿轮箱振动信号频谱特征;运用谐波乘积谱(Harmonic product spectrum,HPS)与边带乘积谱(Sideband product spectrum,SPS)的方法,在噪声干扰以及故障冲击不明显的条件下,准确提取到了仿真信号的故障特征频率。进一步采集不同运行工况、不同故障状态下的行星齿轮箱振动信号,将提取后的故障特征输入到卷积神经网络中进行故障识别,成功获取到齿轮箱的故障信息,证明了该方法在行星齿轮箱故障诊断方面的可行性。

In order to solve the problem that the complex structure and operating conditions of planetary gear box lead to the difficulty of signal fault feature extraction,the frequency spectrum feature of gearbox vibration signal with faults is preliminarily deduced by analyzing the vibration mechanism of planetary gear train.The method of harmonic product spectrum(HPS)and sideband product spectrum(SPS)is used to accurately extract the fault characteristic frequencies of the simulation signals under the condition that the noise interference and fault impact are not obvious.The vibration signals of the planetary gearbox under different operating conditions and different fault states are further collected,and the extracted fault features are input into the convolutional neural network for fault identification.The fault information of the gearbox is obtained successfully,which proves the feasibility of the proposed method in fault diagnosis of the planetary gearbox.