Triplet Loss Guided Adversarial Domain Adaptation for Bearing Fault Diagnosis

Abstract

Recently, deep learning methods are becomingincreasingly popular in the field of fault diagnosis and achieve great success. However, since the rotation speeds and load conditions of rotating machines are subject to change during operations, the distribution of labeled training dataset for intelligent fault diagnosis model is different from the distribution of unlabeled testing dataset, where domain shift occurs. The performance of the fault diagnosis may significantly degrade due to this domain shift problem. Unsupervised domain adaptation has been proposed to alleviate this problem by aligning the distribution between labeled source domain and unlabeled target domain. In this paper, we propose triplet loss guided adversarial domain adaptation method (TLADA) for bearing fault diagnosis by jointly aligning the data-level and class-level distribution. Data-level alignment is achieved using Wasserstein distance-based adversarial approach, and the discrepancy of distributions in feature space is further minimized at class level by the triplet loss. Unlike other center loss-based class-level alignment approaches, which hasto compute the class centers for each class and minimize the distance of same class center from different domain, the proposed TLADA method concatenates 2 mini-batches from source and target domain into a single mini-batch and imposes triplet loss to the whole mini-batch ignoring the domains. Therefore, the overhead of updating the class center is eliminated. The effectiveness of the proposed method is validated on CWRU dataset and Paderborn dataset through extensive transfer fault diagnosis experiments.

Prognostics and Health Management of Industrial Assets: Current Progress and Road Ahead

“…These approaches draw inspiration from the training procedure used by the popular Generative Adversarial Networks (GANs) (Goodfellow et al, 2014) to efficiently align source and target domain features in a common latent-space. Several new techniques (Han et al, 2019a;Wang et al, 2019a;Wang and Liu, 2020) based on this class of DA approaches have been recently proposed in the PHM literature. Other references on DA and TF approaches in the context of fault diagnosis can be found in the recent review works of Li et al (2020) and Zheng Z. et al, 2019.…”

Prediction Consistency Guided Convolutional Neural Networks for Cross-Domain Bearing Fault Diagnosis

“…The Wasserstein Distance guided Multi-Adversarial Networks (WDMANs) [33] employ a multiple-domain critic network to learn the shared feature representation between the source domain and target domain. The Triplet Loss guided Adversarial Domain Adaptation (TLADA) [34] aligns the domain distributions using the Wasserstein distance to match the class distribution by assigning pseudolabels for target samples. Our approach differs from the previous methods in that two mutually reinforced networks rather than a single network are designed in our model to capture the semantic information and estimate the target distribution simultaneously.…”

“…Throughout the experiments, the hyperparameters are empirically set as follows: the prediction consistency weighting parameter is β = 3, the class balance weighting parameter is γ = 0.5, and the confidence threshold of the predicted target labels is θ = 0.96. We adopt the accuracy, precision and recall as our evaluation metrics; these statistics are widely used for model ability assessments in bearing fault diagnosis [33] [34]. Here, we introduce the definition of the confusion matrix for classification, as shown in Table 2.…”

“…In setting II, we focus on bearing fault diagnoses based on vibration signals acquired from different sensor locations. By following the experimental configuration in [34], we select vibration data sampled on the drive end and fan end with the sampling frequency of 12 kHz to construct the source and target datasets. In each domain, 4 bearing conditions, i.e., normal, inner race fault, ball fault and outer race fault, are considered as 4 categories.…”

A Subspace Based Transfer Joint Matching with Laplacian Regularization for Visual Domain Adaptation

“…Indoor-outdoor camera surveillance systems [1,2] are widely used in urban areas, railway stations, airports, smart homes, and supermarkets. These systems play an important role in security management and traffic management [3].…”