Research on an intelligent diagnosis method of mechanical faults for small sample data sets

Abstract

The difficulty of feature extraction and the small sample size are two challenges in the field of mechanical fault diagnosis for a long time. Here we propose an intelligent mechanical fault diagnosis method for scenario with small sample datasets. This method can not only diagnose bearing faults but also gear faults, and has strong generalization performance. We use convolutional neural network to realize automatic feature extraction. Through sliding window scanning, one sample set is expanded to three sub-sample sets with different scales to meet the needs of deep learning training. Three convolutional networks are used to extract the features of the subsets respectively to ensure that their useful features are fully extracted. After feature extraction, the feature is reconstructed through feature splicing. Because of the unique advantages of SVM in dealing with small sample sets, we use SVM to classify the reconstructed features. We use the bearing data set collected by Case Western Reserve University in the United States, the bearing fault data set collected by Xi'an Jiaotong University in China, and the gearbox fault data collected by the University of Connecticut in the United States to conduct experiments. The experimental results show that the accuracy of training, validation and testing of the proposed method on the three data sets all reach 100%. This proves that our method can not only tackle the two challenges, but also has high fault diagnosis accuracy and strong generalization performance. It is hoped that our proposed method can contribute to the development of mechanical fault diagnosis.

Mechanical fault diagnosis based on deep transfer learning: a review

“…According to the need for real-time fault diagnosis of early bearings, Li et al [8] summarized the transfer learning method in dealing with the difficulties of early fault diagnosis of bearings. Zhao et al [9] for the transfer learning method for small samples of bearing fault diagnosis application is summarized and put forward the transfer learning needs sufficient similar samples, otherwise, negative migration may occur, Zhao et al [10] to transfer learning applied to device fault diagnosis, the migration device fault diagnosis is divided into two domain tag sharing device fault diagnosis, source domain contains target domain tag device fault diagnosis and two domain tag device fault diagnosis. The above paper is mainly for migration learning in some working conditions or using some type of neural network under the premise of part of the review, and migration learning working conditions is very complex and uses several neural networks, so it is difficult to understand the efficiency of migration learning algorithm, it is difficult to fully understand the application in the field of fault diagnosis.…”

Mobile user experience classification based on random forest model

“…Among them, customer satisfaction has become an important reflection of the market operation of major operators in terms of customers' network use experience, reflecting the difference between customer expectations and actual perceived product services. The traditional method of solving problems based on customer complaints to improve customer satisfaction is no longer applicable to contemporary society [2,3,4,5] . With the substantial increase in the number of users, the variety of mobile products is becoming more and more abundant, and the demand of customers is becoming higher and higher, so traditional methods have been difficult to effectively improve customer satisfaction [6] .…”