Xiangyu Shan

Xiangyu Shan

Auto-generated profile, based on publicly available information

Xiangyu Shan

Southwest Jiaotong University

Research Focus & Metrics

Academic Career Length

2Years (2025-Present)

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

Representative Publications

View All Publications
1 January 2026 in IEEE Transactions on Transportation Electrification
Verified
Journal Article
Closed access
QZQiao Zhang
Qiao Zhang
XSXiangyu Shan
Xiangyu Shan
XMXiangyu Meng
Xiangyu Meng
ZLZhigang Liu
Zhigang Liu
WGWenxiang Guo
Wenxiang Guo
YZYougang Zhang
Yougang Zhang

22 December 2025 in IEEE Transactions on Transportation Electrification
Verified
Journal Article
Closed access
YLYanming Lu
Yanming Lu
QZQiao Zhang
Qiao Zhang
XMXiangyu Meng
Xiangyu Meng
HYHaonan Yang
Haonan Yang
ZLZhenzu Liu
Zhenzu Liu
ZLZhigang Liu
Zhigang Liu
In electrified railways, the train generates substantial regenerative braking energy (RBE) during braking mode. Deploying wayside hybrid energy storage systems (HESSs) offers a promising solution for recovering and reusing this energy. However, due to the highly fluctuating nature of train power, existing methods based on fixed timescales often struggle to achieve an effective tradeoff between computational efficiency and modeling accuracy. To resolve these limitations, this article proposes a novel operation approach for HESS based on a variable-timebox model. First, a variable-timebox model is constructed, in which the charging and discharging durations of HESS are treated as decision variables. The model simultaneously optimizes RBE recovery and battery capacity degradation by considering the depth of discharge (DoD). Second, to determine the reasonable duration range, a temporal clustering method is proposed. A change point detection (CPD) strategy based on an improved cumulative sum (CUSUM) algorithm is employed to identify characteristics of the traction load. Finally, the effectiveness of the proposed method is validated using real-world data. The simulation results demonstrate that the proposed method reduces computation time by more than 60% while maintaining modeling accuracy.