Yan Liu

Dr. Yan Liu

orcid

Research Focus & Metrics

Academic Career Length

4Years (2023-Present)

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

Representative Publications

View All Publications
21 November 2025 in Engineering Applications of Artificial Intelligence
Verified
Journal Article
Closed access
JHJialong He
Jialong He
ZMZhenbiao Ma
Zhenbiao Ma
YLYan Liu
Yan Liu
ZYZhaojun Yang
Zhaojun Yang

15 May 2025 in Measurement Science and Technology
Verified
Journal Article
Closed access
JHJialong He
Jialong He
ZMZhenbiao Ma
Zhenbiao Ma
YLYan Liu
Yan Liu
ZYZhaojun Yang
Zhaojun Yang
Remaining useful life (RUL) prediction is critical for ensuring the safe and efficient operation of rotating machinery, which plays a vital role in industrial production. However, the traditional transformer model, relying on a simple linear output in its final layer, is limited in capturing nonlinear relationships. The nonlinear Wiener process (WP) model, despite its strengths in characterizing stochastic degradation, fails to fully capture the intricate time-series features in equipment operation data. To address these challenges, this paper introduces an improved method for RUL prediction which combines transformer and Kolmogorov–Arnold network (KAN) and utilizes WP for uncertainty quantification. The transformer can efficiently capture the long-term dependencies in equipment degradation data through its powerful sequential modeling capability and global attention mechanism, and its combination with the KAN further enhances the flexibility and accuracy of feature extraction. In this paper, the transformer-KAN model is proposed as a drift function for WP. Meanwhile, this paper utilizes the first hit time to derive an approximate expression for the RUL probability density function and estimate the drift and diffusion coefficients of the transformer-KAN-WP model. The method is validated on the bearing dataset and the tool holder power head degradation data and compared and analyzed with other commonly used methods, which proves the effectiveness of the method in improving the prediction accuracy.

1 January 2023 in International Journal of Intelligent Systems
Verified
Journal Article
Open access
YLYan Liu
Yan Liu
ZYZhaojun Yang
Zhaojun Yang
JHJialong He
Jialong He
LYLijuan Yu
Lijuan Yu
YZYuan Zhong
Yuan Zhong
Intuitionistic cubic fuzzy sets (ICFSs), as a hybrid fuzzy set consisting of interval‐valued fuzzy sets and intuitionistic fuzzy sets, can deal with the uncertainty of information in decision problems more comprehensively. Muirhead mean (MM) operators can deal with the correlation between attributes in decision problems more realistically. However, there has been no research on MM operators under intuitionistic cubic fuzzy sets so far. In this study, some new operators are proposed under the intuitionistic cubic fuzzy set. It contains the intuitionistic cubic fuzzy MM (ICFMM) operator, intuitionistic cubic fuzzy weighted MM (ICFWMM) operator, intuitionistic cubic fuzzy dual MM (ICFDMM) operator, and intuitionistic cubic fuzzy dual weighted MM (ICFDWMM) operator. Their proof procedure, properties, and proof of properties are given. Two MCDM decision models based on ICFWMM and ICFDWMM operators are proposed, and the proposed decision models are applied to supplier selection, and the better reliability and accuracy of the proposed decision methods in this study are verified through comparative analysis with existing methods.

4 September 2025 in IEEE Transactions on Instrumentation and Measurement
Verified
Journal Article
Closed access
JHJialong He
Jialong He
JGJichao Guo
Jichao Guo
Liming ZhouLiming Zhou
Liming Zhou
YLYan Liu
Yan Liu
Accurate prediction of the remaining useful life (RUL) constitutes the core task of predictive maintenance, and the selection of an appropriate degradation model is pivotal to enhancing the accuracy of RUL prediction. However, deep learning models cannot characterize RUL uncertainty, and traditional stochastic process degradation models are challenging in depicting the long-range dependence (LRD) of the degradation process, adversely impacting the accuracy and credibility of RUL prediction. To address these challenges, this paper unveils a synergistic fractional Brownian motion (TCN-KAN-FBM) prediction framework using temporal convolutional networks and Kolmogorov-Arnold Networks for robust device RUL prognostication. The TCN-KAN module is designed to realize RUL prediction. The TCN-KAN module captures temporal features of degraded data and adaptively learns degradation knowledge for point estimation prediction. Complementing this, the FBM module then masterfully constructs the probability distribution of the prediction results based on its long-range dependence and self-similarity, thus realizing RUL prediction’s uncertainty quantification (UQ). The effectiveness of the proposed method is confirmed by practical examples of rolling bearings and two sets of servo tool holder power head systems under different operating conditions.

18 November 2025 in IEEE Internet of Things Journal
Verified
Journal Article
Closed access
CQChenhui Qian
Chenhui Qian
ZYZhaojun Yang
Zhaojun Yang
JHJialong He
Jialong He
YLYan Liu
Yan Liu
MYMingdong Yang
Mingdong Yang
SLShaoyang Liu
Shaoyang Liu
Although meta-learning models have achieved some success in few-shot cross-domain diagnosis tasks, they still face three major challenges: insufficient feature extraction capability, an uncontrollable domain adaptation process, and low discriminability of class features. To address these problems, a few-shot cross-domain fault diagnosis method for complex mechanical equipment based on multi-angle Taylor prototypes (MATM) is proposed. First, after constructing the dynamic graph autoencoder, the adjacency matrix is updated during the training process to capture the dynamic variations in fault data, which enhances the model’s ability to learn key features. Next, a Gaussian distribution-guided domain adaptation module is constructed, which imposes a Gaussian prior constraint on the latent space during the domain adaptation process to guide the optimization direction of domain-invariant features, thereby promoting the alignment of feature distributions across different domains. Finally, a multi-angle Taylor prototype metric module is constructed to perform nonlinear expansion of supporting features in high-dimensional space, optimizing the quality of class prototype representations from multiple perspectives to more accurately capture the class relationships across different domains. Experimental results on reducer datasets show that, compared to the latest methods, the proposed method improves the average classification accuracy by 1.98% across all diagnostic tasks, demonstrating superior performance in small-sample cross-domain fault diagnosis tasks.

2 January 2026 in Quality and Reliability Engineering International
Verified
Journal Article
Closed access
JHJialong He
Jialong He
TYTianshuo Yu
Tianshuo Yu
YLYan Liu
Yan Liu
CMChi Ma
Chi Ma
XGXiufeng Gao
Xiufeng Gao