Xinyu Chang

Xinyu Chang

Auto-generated profile, based on publicly available information

Xinyu Chang

School of Electrical and Electronic Engineering, Hubei University of Technology, Wuhan, China

Research Focus & Metrics

Academic Career Length

3Years (2024-Present)

(Calculated once Scilit publication data is synced: current year − year of first publication + 1)

Representative Publications

View All Publications
24 October 2024 in IEEE Sensors Journal
Verified
Journal Article
Closed access
LZLiang Zeng
Liang Zeng
XCXinyu Chang
Xinyu Chang
JCJia Chen
Jia Chen
SWShanshan Wang
Shanshan Wang
Deep learning-based methods have made remarkable achievements in rolling bearing fault diagnosis in recent years. Nevertheless, due to the diversity and complexity of rolling bearing working conditions, how to generalize deep learning models trained under limited conditions to unseen working conditions has become a popular issue in current research. The existing methods based on domain generalization usually need to train the model using a significant quantity of labeled data under multiple known working conditions to enhance its generalization capabilities under unseen working conditions. However, the acquisition of adequately labeled samples is a time-consuming and laborious task. Therefore, this paper proposes an auxiliary branch semi-supervised domain generalization network (ABSDGN). ABSDGN employs a joint learning strategy of the auxiliary and main branches. The auxiliary branch generates high-confidence pseudo-labels for the unlabeled source-domain data using the labeled source-domain data. The main branch leverages both real and pseudo labels to learn domain-invariant knowledge. Meanwhile, a quadratic neuron-based convolution (QConv) is introduced to take advantage of its powerful nonlinear and higher-order feature representation capabilities in enhancing the performance and generalization of the model in complicated situations. In addition, a weight decomposition method based on domain labels is proposed to decompose the main branch classifier into a generic feature and feature-specific classifier to better explore the universal knowledge among different domains. The experimental results show that the proposed model has better diagnostic accuracy and stability than the most advanced semi-supervised domain generalization method on two-bearing datasets.

18 March 2025 in IEEE Sensors Journal
Verified
Journal Article
Closed access
SWShanshan Wang
Shanshan Wang
WHWenkang Han
Wenkang Han
JJJunjie Jian
Junjie Jian
XCXinyu Chang
Xinyu Chang
LZLiang Zeng
Liang Zeng
In industrial applications, intelligent fault diagnosis technology is rapidly evolving, with deep learning-based intelligent diagnosis methods proving effective in managing and maintaining equipment. However, the variability of working conditions for mechanical equipment in actual industrial settings often presents challenges. The data collected by sensors from different working conditions or machines may exhibit significant differences in distribution. In addition, it is difficult to collect a large number of labeled samples. This paper introduces a domain-adaptive meta-learning method with multilayer convolution alattention, named (MCA-DAML), designed for cross-domain bearing fault diagnosis. The proposed approach involves taking the source domain data and unlabeled target domain data as inputs. Through adversarial domain adaptation, the model is trained to simultaneously minimize the source domain task loss and maximize the confusion error of the domain discriminator. This dual optimization strategy encourages the model to learn shared feature representations that are effective across different domains. Multi-layer convolutional attention modules are used to enhance the feature extraction capabilities of the model and suppress redundant features, which are analyzed by the prototype network for proximity to established fault prototypes, ultimately achieving accurate fault classification. Evaluated using three bearing vibration datasets without labeling the target domain samples, the average accuracy achieved was 99.86%, 96.51%, and 94.24% when performing the three cross-domain cases within the same machine and between different machines, respectively. The experimental results validate the superior performance of our method relative to other diagnostic methods.