Based on the characteristics of non-linear, non-stationary, and strong noise of rolling bearing faults, aiming at the problems that traditional data-driven fault diagnosis methods rely on manual feature extraction on the one hand, it is difficult to adaptively extract effective feature components, on the other hand, they fail to fully mine the effective timing features of fault data and do not have the ability to adaptively extract dynamic information. An intelligent fault diagnosis method combining variational modal decomposition (VMD) with convolution neural network (CNN) and long short term memory (LSTM) network is proposed. First, the original bearing vibration signal is decomposed into variable modes to obtain the modal components containing fault feature information, and the local modal components with prominent features are identified using the instantaneous center frequency method and formed into a two-dimensional feature matrix with the original data as the input of (CNN), and the fault feature information is extracted implicitly and adaptively using CNN network as the input of LSTM layer, then the LSTM layer is used to learn the features. Finally, the output layer of the target model is used to realize the pattern recognition of multiple faults of the bearing using Softmax function to complete the fault diagnosis. The experimental results show that the method improves the accuracy of the diagnosis results and the shortcomings of the traditional diagnosis methods.
Title : Application of VMD Combined with CNN and LSTM in Motor Bearing Fault Name : Ran Song