Rotating machinery fault diagnosis based on optimized Hilbert curve images and a novel bi-channel CNN with attention mechanism

Abstract

Deep learning methods have been widely investigated in machinery fault diagnosis owing to their powerful feature learning capability. However, high accuracy is hard to achieve due to the limited fault information in a single domain when the data volume is small. In this paper, an optimized Hilbert Curve (OHC) method is developed, which can generate a novel domain to highlight the fault impulses of vibration signals. To fully mine the fault information, a bidirectional-channel convolutional neural network with an attention mechanism (BCCNN-AM) is further proposed, in which two channels are constructed and a transmission channel selection is conducted by a novel improved convolutional block attention module (ICBAM).
First, the OHC images and the time-frequency representations are obtained by OHC and wavelet transform respectively. Second, the two types of representations are fed into the channels respectively for feature learning. Finally, the learned features are allocated to different attention mechanism for feature fusion and classification. The proposed method is evaluated via the datasets of rolling bearings and planetary gearboxes, and results show that it outperforms the comparison methods.

Multiscale permutation entropy gray image coding method and its application in bearing fault diagnosis

“…The MPEGIC method proposed in this paper uses a CNN with a parameter size of only 1710058, achieving the highest accuracy rate of 99.57%. [35], added a Channel Segmentation Mechanism (CSM) on top of Resnet34, which requires more computer memory resources, resulting in an accuracy rate of 99.4%, still 0.17% lower than that of this paper.…”

“…To realize bearing fault diagnostics [34], combines (MTF) and multi-scale Runge-Kutta residual network (MRKRA-Net). [35], suggests using CNN and the optimal Hilbert curve (OHC) approach for bearing fault identification. [36], transforms one-dimensional vibration signals into two-dimensional timefrequency pictures using the generalized S-transform (GST).Although these approaches are capable of extracting features from vibration signals and detecting faults, they still have the following drawbacks when used to diagnose bearing faults.…”

Rotating Machinery Fault Diagnosis Under Time-Varying Speeds: A Review

“…Deep learning-enabled diagnosis means that the effective features are learned from the raw signals or the signals through simple preprocessing, such as transforming a 1-D data to a 2-D TFR, automatically by deep neural networks, such as deep autoencoders [187] , deep belief networks [188] and convolutional neural networks (CNNs) [189] , [190] , [191] . Recently, some researchers have attempted to construct deep diagnosis models to implement fault recognition under time-varying speeds.…”

A review on convolutional neural network in rolling bearing fault diagnosis

“…On the HHT, Zhao et al [60] proposed the a new interpretable denoising layer based on reproducing kernel Hilbert space as the first layer of the standard neural network, aiming to combine traditional signal processing techniques with physical interpretation and network modeling strategies and parameter adaptation. Sun et al [61] proposed an optimized Hilbert curve (OHC) method that can generate a new domain to highlight fault pulses in vibration signals. Firstly, The time-frequency representations were obtained through OHC and wavelet transform.…”