A VMD-AM-LSTM-based method for health condition assessment of liner-type self-lubricating joint bearings

The liner type self-lubricating joint bearings are simple in structure and have large load bearing and impact resistance, which are widely used in aerospace, automotive, construction machinery and many other fields. The accurate assessment of their health status can provide a decision basis for the repair and maintenance of the equipment, which can effectively reduce the safety accidents and economic losses. Considering multiple features, such as the friction torque, wear amount and friction temperature and so on, a VMD-AM-LSTM-based method for health condition assessment of liner-type self-lubricating joint bearings is proposed, which is able to predict the degradation tendency of the liner type self-lubricating joint bearing more accurately in order to grasp its health condition. Firstly, a set of fatigue wear tests of liner-type self-lubricating joint bearings under multiple working conditions is carried out by using a comprehensive evaluation testing machine for joint bearings to collect signals such as friction torque, wear and temperature during the degradation of the bearings. Secondly, the VMD method is used to extract features from the collected signals in order to obtain the feature components that can characterize the bearing degradation trend. Then the feature components are normalized and input to the VMD-AM-LSTM model for training and testing. Finally, the Bayesian optimization method is used to optimize the hyperparameters to further improve the prediction accuracy of the model. A series of comparative experimental results show that the proposed method has higher prediction accuracy for the health condition of liner type self-lubricating joint bearings under different working conditions.