An Intelligent Fault Diagnosis Approach Considering the Elimination of the Weight Matrix Multi-Correlation

Abstract

Faults in bearings and gearboxes, which are widely used in rotating machines, can lead to heavy investment and productivity loss. Accordingly, a fault diagnosis system is necessary to ensure a high-performance transmission. However, as mechanical fault diagnosis enters the era of big data, it can be difficult to apply traditional fault diagnosis methods because of the massive computation cost and excessive reliance on human labor. Meanwhile, unsupervised learning has been shown to have excellent performance in processing machanical big data. In this paper, an unsupervised learning method known as sparse filtering is applied, the multi-correlation of a weight matrix is investigated, and a method that is more suitable for the feature extraction of signals is proposed. The main contribution of our work is the modification of original method. First, to understand the non-monotonicity testing accuracies of the original method, the physical interpretation of input dimensions is studied. Second, using the physical interpretation, an overfitting phenomenon is discovered and examined. Third, to reduce the overfitting phenomenon, a method which eliminates the multi-correlation of the weight matrix is proposed. Finally, bearing and gear datasets are employed to verify the effectiveness of the proposed method; experimental results show that the proposed method can achieve a superior performance in comparison to the original sparse filtering model.

Adaptive Cross-Domain Feature Extraction Method and Its Application on Machinery Intelligent Fault Diagnosis Under Different Working Conditions

“…So in order to comprehensively check the health condition of the machinery, a large amount of signals are obtained after the long-time monitoring, which also brings great difficulties to mechanical fault diagnosis. Therefore, various intelligent algorithms have been proposed for fault diagnosis, such as Artificial Neural Networks (ANN) [6], [7], Autoencoders [8], [9], Restricted Boltzmann Machine (RBM) [10], Convolutional Neural Networks (CNN) [11], [12], Sparse Filtering [13], [14] and k-Nearest Neighbor [15].…”

A Novel Data-Driven Fault Feature Separation Method and Its Application on Intelligent Fault Diagnosis Under Variable Working Conditions

“…Machine learning which is regarded as a data-driven model, depends less human knowledge and is quite suitable to deal with mechanical big data, has drawn wide attention in fault diagnosis field, such as Artificial Neural Networks (ANN) [14,15], Autoencoders (AE) [16,17], Restricted Boltzmann Machine (RBM) [18], Convolutional Neural Networks (CNN) [19,20], Sparse Filtering [21,22] and k-Nearest Neighbor [23]. For the fault diagnosis under different working conditions, deep transfer learning, as a branch of deep learning, has been employed.…”

Research and Application of Regularized Sparse Filtering Model for Intelligent Fault Diagnosis Under Large Speed Fluctuation

“…Yang et al [20] used SF to extract sparse features directly from raw timedomain signals and classified health status using traditional support vector machines based on improved particle swarm optimization. An et al [21] solved the over fitting problem in bearing fault diagnosis by removing multi-correlation operations in the weight matrix of SF.…”