基于FMD-DO-DCFS的自适应特征模式轴承故障诊断OA

Aiming at the issues of low diagnostic accuracy and difficulty in adapting to variable working conditions in traditional intelligent fault diagnosis methods for rolling bearings,an adaptive eigenmode bearing fault diagnosis method based on eigenmode decomposition,dandelion optimization algorithm,and deep convolutional fuzzy system was proposed to enhance diagnostic accuracy.Firstly,the fault signal of the rolling bearing was filtered via eigenmode decomposition to extract the denoised vibration signal,and time-domain calculations were performed on the collected bearing data to construct the fault feature vector.Secondly,the dandelion optimization algorithm was employed to rank and select the fault features which have the largest informative quantity,and then the selected features were fed into the pre-established deep convolutional fuzzy system model for training and testing,yielding the classification results of bearing fault diagnosis.Finally,a comparative experiment was conducted according to the bearing data set to evaluate the proposed method against five other approaches.The results demonstrate that the proposed method can achieve more accurate fault diagnosis for rolling bearings under high-noise conditions.