Research On Heterogeneous Features Fusion And Online Imbalance Classification For Bearing Fault Diagnosis

Rolling bearings are one of the most commonly used parts in rotating machinery.The condition when machinery works have a critic influence to the safety of the equipment and even the whole production line.Therefore,it is of great theoretical and practical significance to study the condition monitoring and fault diagnosis technology of rolling bearing to avoid the occurrence of severe accident and reform the maintenance system.In this paper,we analyzed the characteristics of rolling bearing data and proposed some algorithms to prove the performance of bearing fault diagnosis.The main contents are summarized as follows:(1)To address the problem of fault diagnosis with no universal features or with relatively simple form of feature,a heterogeneous feature fusion method based on the feature correlation among groups is proposed for the bearing fault diagnosis.The heterogeneous feature extracted by different methods has the complementary effect.The proposed method combined all the heterogeneous features into a joint feature set.Particle group method is used to optimize the grouping of these features based on feature correlation by maximizing intra-group feature correlation and minimizing the inter-group feature discrimination at the same time.Then,the feature of each group is selected by the wrapper algorithm as the final selected fusion feature.This method effectively combines heterogeneous features and removes the redundancy between features.Finally,the simulation results are carried out on IMS and CWRU bearing data.(2)The bearing data tend to be online imbalanced,which means,the number of fault data is much less than the normal data while they are all collected in online sequential way.Suffering from this problem,many traditional diagnosis methods will get low accuracy of fault data which acts as the minority class in the collected bearing data.To address this problem,an online sequential prediction method for imbalanced fault diagnosis problem is proposed based on extreme learning machine.This method introduces the principal curve and granulation division to simulate the flow distribution and overall distribution characteristics of fault data,respectively.Then a confident over-sampling and under-sampling process is proposed to establish the initial offline diagnosis model.In online stage,the obtained granules and principal curves are rebuilt on the bearing data which are arrived in sequence,and after the over-sampling and under-sampling process,the balanced sample set is formed to update the diagnosis model dynamically.A theoretical analysis is provided and proves that,even existing information loss,the proposed method has lower bound of the model reliability.Simulation experiments are conducted on IMS bearing data and CWRU bearing data.The comparative results demonstrate that the proposed method can improve the fault diagnosis accuracy with better effectiveness and robustness than other algorithms.