Guiyang Hu

Guiyang Hu

Auto-generated profile, based on publicly available information

Guiyang Hu

School of Electrical Engineering, Southwest Jiaotong University, Chengdu, China

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
15 May 2024 in IEEE Transactions on Industrial Informatics
Verified
Journal Article
Closed access
HWHui Wang
Hui Wang
ZLZhigang Liu
Zhigang Liu
GHGuiyang Hu
Guiyang Hu
XWXufan Wang
Xufan Wang
ZHZhiwei Han
Zhiwei Han
Previous reinforcement learning (RL) methods suffer significant performance degradation or collapse when deployed to the real world due to the huge sim-real gap. This article proposes a hybrid offline-and-online meta-RL (HOMRL) algorithm that leverages prior task experience to learn and adapt to new pantograph active control tasks in real-world applications. The policy learning process consists of three phases: offline meta-policy pretraining, online adaptation, and fine-tuning. First, we construct an offline meta-RL approach that learns from the massive and heterogeneous static training datasets, eliminating online interaction's high cost and hazard. Second, we combine context-based meta-RL with online fine-tuning to generalize to challenging tasks, while high safety and success rates are critical in railway applications. Finally, the proposed environment-sensitive task encoder (TE) and well-trained agent can adapt to new tasks quickly and efficiently, even in unseen tasks and nonstationary environments. If the new task is similar to the prior data, the contextual meta-learner adapts immediately. If it is too different, it gradually adapts through fine-tuning.

17 September 2024 in IEEE Transactions on Transportation Electrification
Verified
Journal Article
Closed access
XMXiangyu Meng
Xiangyu Meng
GHGuiyang Hu
Guiyang Hu
ZLZhigang Liu
Zhigang Liu
HWHui Wang
Hui Wang
GZGuinan Zhang
Guinan Zhang
Mahdieh S. SadabadiMahdieh S. Sadabadi
Mahdieh S. Sadabadi
The double-sided power supply railway system increases the simultaneous operation of vehicles on the grid, potentially causing system instability and oscillation overvoltage issues. As vehicles frequently switch operating points during operation, it is essential to analyze system stability across a wide range of conditions. Therefore, accurately identifying the black-box impedance of vehicle converters at multiple operating points is crucial for studying railway vehicle-grid system stability. However, traditional impedance identification methods require extensive data and lack interpretability, leading to significant computational and data burdens. This study introduces an interpretable residual feedforward neural network (ResFNN) combined with SHapley Additive exPlanations for training vehicle impedance models, reducing data requirements while maintaining accuracy. Additionally, a component connection method is proposed for deriving the impedance matrix of a multi-vehicle railway system under the double-sided feeding mode. This method incorporates the dynamic mobility of vehicles and their positional distribution, and it utilizes the ResFNN to identify impedance for stability analysis. Real operational data from actual railway lines is used as case study to analyze the stability of the double-sided power supply railway system. The results demonstrate that this approach accurately assesses both low-frequency and high-frequency instability issues.

27 December 2023 in IEEE Transactions on Power Electronics
Verified
Journal Article
Closed access
XMXiangyu Meng
Xiangyu Meng
QZQiao Zhang
Qiao Zhang
ZLZhigang Liu
Zhigang Liu
GHGuiyang Hu
Guiyang Hu
FLFang Liu
Fang Liu
GZGuinan Zhang
Guinan Zhang
The electrical incompatibility between vehicles and traction network in railway system can result in system instability and oscillation overvoltage issues. To analyze the system stability, impedance-based frequency-domain methods are commonly used. However, the current impedance-based modeling methods face challenges in practical implementation due to the requirement of precise analytical models and detailed internal parameters for all vehicles. Moreover, multiple vehicles operate simultaneously in railway systems, each with different operating conditions and internal parameters, thereby influencing system stability to different extents. Therefore, it is crucial to accurately identify the critical vehicles to prevent resonance accidents. To address these challenges, a component connection-based modeling approach for the railway vehicle-grid system is proposed, which only requires the measured impedance results without the internal information of vehicles. In addition, a multilevel sensitivity analysis method is introduced to quantitatively identify the critical vehicles and internal parameters that influence system stability, which outperforms traditional sensitivity analysis methods in computational complexity. Furthermore, a system-level electrical compatibility test process for the railway vehicle-grid system is provided, incorporating the proposed stability and sensitivity analysis methods. Finally, case studies based on the real-world train schedule of a multivehicle-accessed railway vehicle-grid system are designed to verify the correctness of the proposed method.

30 September 2024 in IEEE Transactions on Instrumentation and Measurement
Verified
Journal Article
Closed access
XWXufan Wang
Xufan Wang
ZLZhigang Liu
Zhigang Liu
Yang SongYang Song
Yang Song
GHGuiyang Hu
Guiyang Hu
SZSheng Zhou
Sheng Zhou
LCLong Chen
Long Chen
In high-speed railways, the pantograph serves as the sole apparatus for collecting electric energy, directly impacting the operational safety of high-speed trains through its structural stability. As train speed increases, the interaction between the pantograph and the catenary becomes more pronounced, potentially compromising the structural safety of the pantograph during operation. Previous studies have not fully accounted for the significant effects of large fluctuations in the high-speed pantograph-catenary contact force on the pantograph structure. This paper addresses the issue by establishing a comprehensive finite element model of the pantograph that incorporates parametric structural dimensions. The validation of the model is confirmed through lumped mass parameter identification and modal analysis tests. Additionally, a dynamic coupling model of the pantograph-catenary system is developed to assess the load impact on the pantograph. By applying the load, the dynamic deformation and stress of the pantograph are analyzed to identify critical structural locations during high-speed operation. The findings reveal that the structural strength of the pantograph at 400 km/h falls below the permissible threshold. Consequently, a neural network-based optimization approach is proposed to minimize the maximum stress and deformation of the pantograph at this speed. The pantograph’s structural performance response surface is constructed using sensitivity analysis and a radial basis neural network. Finally, a multi-objective genetic algorithm (MOGA) is employed to search the response surface and achieve optimal results.

28 June 2024 in IEEE Transactions on Transportation Electrification
Verified
Journal Article
Closed access
GHGuiyang Hu
Guiyang Hu
XMXiangyu Meng
Xiangyu Meng
XWXiaokang Wang
Xiaokang Wang
ZLZhigang Liu
Zhigang Liu
As many Electric Multiple Units (EMUs) integrate into the traction network, low-frequency oscillation issues may occur in railway systems. Impedance-based frequency-domain stability analysis is a common method used to analyze the stability of vehicle-grid systems. However, the vehicle-grid system is a complex single-phase system with numerous nonlinear components, making it difficult to establish accurate and online-analyzable impedance models. To address these challenges, this paper proposes the Residue Feedforward Neural Network (ResFNN) suitable for EMU impedance identification. The ResFNN integrates residual connections into the deep Feedforward Neural Network (FNN), which can improve the model prediction accuracy and get rid of the complex model derivation process. To explain the contribution of the input parameter to the model, the SHapley Additive exPlanation (SHAP) method is adopted in this paper. Further, according to the above contributions, the paper proposes an unequal step size data collection method to optimize the operating point parameter step size and obtain a small but high-quality dataset. Then, combined with the vehicle-side and grid-side impedance models, this paper can realize online vehicle-grid system stability analysis. Finally, actual case studies are conducted with multiple vehicles connected to the vehicle grid system to validate the feasibility and accuracy of online stability analysis.