Research On Fault Diagnosis Method Of Rolling Bearing Based On Generative Adversarial Networks

With the continuous development of modern society and the progress of science and technology,industrial production is increasingly intelligent,so higher requirements are put forward for the safety and precision of mechanical equipment.Guaranteeing the safety and stability of machinery equipment in real industrial scenarios can not only reduce the production and maintenance costs of enterprises,but also avoid the occurrence of major safety accidents.Therefore,the research of equipment condition monitoring and fault diagnosis has important practical significance and application value.Based on the theory of generative adversarial networks,this paper proposes two new rolling bearing fault diagnosis methods to solve the data imbalance and data distribution discrepancy problem in real industrial production,which are verified by experiments on the benchmark dataset and the measured dataset of fault diagnosis.The experiment results on the two datasets show that the proposed methods in this paper can not only significantly improve the diagnosis accuracy under the data imbalance and data distribution discrepancy conditions,but they have better stability and generalization ability compared with other methods.The main contents of this paper are as follows:(1)In most time of industrial processes,rolling bearing usually works in normal state and the fault data are hard to collect.The amount of data acquired in normal state is far more than that in faulty state.Therefore,there is an imbalance between normal samples and fault samples,which leads to the performance deterioration of fault diagnosis markedly.To solve this problem,this paper presents a novel imbalanced fault diagnosis method based on the enhanced generative adversarial networks(GAN).By artificially generating fake samples,the proposed method can mitigate the loss caused by the lack of real fault data.Specifically,in order to improve the quality of generated samples,a new discriminator is designed using spectrum normalization strategy and a two time-scale update rule method is used to stabilize the training process of GAN.Then,an enhanced Wasserstein GAN with gradient penalty is developed to generate high-quality synthetic samples for the fault samples set.Finally,a deep convolutional classifier is constructed to carry out fault classification.Experiments are conducted on the imbalanced dataset of rolling bearing,and results demonstrate that the proposed method can effectively improve the fault classification ability with imbalance data.(2)To reduce the domain shift between training and testing data in rotary machinery,a novel transfer learning method called singular value penalization adversarial network is proposed for fault diagnosis of rolling bearings,which first constructs a feature extractor to learn the shared feature representations between the source and target domains automatically,and then plays an adversarial training game between feature extractor and domain discriminator to align the domain shift and achieve domain adaptation.In particular,a new singular value penalization term is proposed and integrated into the domain adversarial network,which balances the distributions of largest singular values and boosts the discriminability of feature representations.The extensive transfer diagnosis experiments on two rolling bearing datasets validate the effectiveness and robustness of the proposed method,and the results show that our approach is superior to other methods under unlabeled data.