Fault Diagnosis Method for Rolling Bearing Based on Sparse Principal Subspace Discriminant Analysis

Abstract

Rolling bearings are omnipresent parts in industrial fields. To comprehensively reflect the status of rolling bearing and improve the classification accuracy, fusion information is widely used in various studies, which may result in high dimensionality, redundancy information of dataset, and time consumption. Thus, it is of crucial significance in extracting optimal features from high-dimensional and redundant feature space for classification. In this study, a fault diagnosis of rolling bearings model based on sparse principal subspace discriminant analysis is proposed. It extracts sparse discrimination information, meanwhile preserving the main energy of original dataset, and the sparse regularization term and sparse error term constrained by l2,1-norm are introduced to improve the performance of feature extraction and the robustness to noise and outliers. The multi-domain feature space involved a time domain, frequency domain, and time-frequency domain is first derived from the original vibration signals. Then, the intrinsic geometric features extracted by sparse principal subspace discriminant analysis are fed into a support vector machine classifier to recognize different operating conditions of bearings. The experimental results demonstrated that the feasibility and effectiveness of the proposed fault diagnosis model based on a sparse principal subspace discriminant analysis algorithm can achieve higher recognition accuracy than fisher discriminant analysis and its extensions, and it is relatively insensitive to the impact of noise and outliers owing to the sparse property.

Prediction of Compressive Strength Loss of Normal Concrete after Exposure to High Temperature

“…X iA ) T . The particle is then updated by each iteration of the individual and extreme global values, with the velocity and position defined in Equations ( 2) and (3) [90].…”

“…In this paper, particle number n was set as 20, learning factors c 1 and c 2 were kept the same as 2 and momentum coefficients r 1 and r 2 were kept the same as 0.8 [90]. Inertia weight ω was assigned to 0.8 after several trials.…”

Vibration characterization of rolling bearings with compound fault features under multiple interference factors

“…At present, the bearing fault diagnosis technology has been widely developed, but there are still problems in the research of bearing fault diagnosis, such as insufficient researches on fault symptoms and fault mechanisms, however, the fault mechanism research is an important foundation and basis for the fault prediction and safety guarantee of key equipment [7].On the basis of mechanical vibration theory, SKF of Sweden and Franklin Institute of the United States [8] cooperated to study the vibration of ball bearings, and put forward a free circular bending vibration model of bearing vibration; Li Yuqi et al [9] analyzed different initial deformation and bearing clearance by establishing a nonlinear rotor system dynamics model and solving the equations of motion of the rotor system by the Newmark-method, and then analyzed the impact of different initial deformation and bearing clearance on the nonlinear vibration of the rotor system. on the nonlinear vibration characteristics of the rotor system; Zhu Yongsheng et al [10] used the Lempel-Ziv method to quantify the bearing vibration signals and established a dynamics model containing six degrees of freedom to study the effect of compound faults on the vibration response of the bearings; Yuan et al [11] constructed a new multi-body dynamics model of the bearing system to study the evolution of compound defects and analyzed the effect of defect evolution on vibration response characteristics; Pandya et al [12] used an adaptive algorithm of wavelet decomposition and Hilbert transform to extract bearing fault feature components from vibration signals, and investigated the nonlinear dynamic response caused by compound faults; Luo Jianing et al [13,14] showed that vibration features obtained from the simulation model can provide effective sample data for the study of bearing fault diagnosis methods, and can be used as a supplement to the fault dataset (S1 Data).…”

Fault diagnosis of rolling bearing based on convolutional neural network and transfer learning

“…In recent years, deep learning and machine learning have been developed continuously, and their outstanding feature extraction and data processing capabilities are widely used in rolling bearing fault diagnosis. Zhou et al [6] combined the vibration signal into image signal with structural reparameterization technique to achieve rolling bearing fault diagnosis. Liu et al [7] decomposed the original bearing signal into multiple multiscale signals by signal decomposition and transformation, and the multiscale signals were used as the corresponding input channels, and the outputs of parallel subneural networks were connected into one channel as the input of the fully connected layer for classification.…”