基于相关向量机的齿轮传动系统故障诊断

Fault Diagnosis of Gear Transmission System Based on Correlation Vector Machine

针对行星齿轮和轴承在实际应用中故障诊断问题,提出了基于相关向量机(RVM)和变分模态分解(VMD)的行星传动系统故障诊断方法。首先,在行星传动系统实验台上进行了6种工况的模拟实验,包括正常状态、行星齿轮断齿、齿面磨损、轴承滚动体缺失、“行星轮断齿+齿面磨损”和“行星轮断齿+轴承滚动体缺失”耦合故障状态;其次,对采集到的振动信号进行VMD分解得到若干个本征模态分量(IMF),通过计算各分量的能量、峭度及峰峰值组成故障特征向量;最后,通过对提取的故障特征进行训练和测试,从而识别区分出单一故障和耦合故障。结果表明:该方法在单一故障识别中基本可完全准确诊断,在耦合故障诊断中也表现较好。研究可为工程实践中行星齿轮箱故障诊断提供理论依据。

Aiming at the problem of fault diagnosis of planetary gear and bearing in practical application,a fault diagnosis method of planetary transmission system based on correlation vector machine(RVM)and variational mode decomposition(VMD)is proposed.Firstly,the simulation experiment of 6 working conditions was carried out on the test platform of planetary drive system,including the normal state,planetary gear broken tooth,tooth surface wear,bearing rolling body loss,"planetary gear broken tooth+tooth surface wear"and"planetary gear broken tooth+bearing rolling body loss"coupling fault state.Secondly,several eigenmode components(IMF)are obtained by VMD decomposition of the collected vibration signals,and fault feature vectors are formed by calculating the energy,kurtosis and peak-to-peak values of each component.Finally,single fault and coupling fault can be distinguished by training and testing the extracted fault features.The results show that the method can be completely and accurately diagnosed in single fault identification,and it also performs well in coupling fault diagnosis.The study can provide theoretical basis for fault diagnosis of planetary gearbox in engineering practice.