To address the braking modeling challenges in medium-low speed maglev trains caused by electric braking response delays and discrete hydraulic compensation in conventional single-mass models, a hybrid TCN-Transformer architecture-based braking model was proposed. A three-stage preprocessing framework was developed, comprising: eddy-current speed measurement data imputation, operational state decomposition, and multi-scale window feature generation. The local temporal pattern extraction capability of TCN (temporal convolutional networks) was integrated with the global dynamic correlation modeling advantages of Transformer, establishing a predictive framework for maglev train braking characteristics. Experimental results demonstrated that MAE (mean absolute error) of 1.114 km/h for 50-step prediction was achieved, which represented a 7.2% improvement over standalone Transformer models. Field validation using in-service track measurement datasets demonstrated that the braking response time was 22.1 s earlier than that of conventional dynamic models, overspeed fluctuations (>0.5 km/h) in maximum speed-restricted segments were completely eliminated, and the terminal positioning error was reduced by 32.4%. It was demonstrated that the hybrid architecture effectively resolved nonlinear modeling challenges in electro-hydraulic hybrid braking dynamic compensation through multi-scale feature fusion, providing a real-time predictive solution for intelligent braking systems in maglev transportation.
WANG Guo, SHI Kai, MIN Yong-zhi, LÜ Wei-xi, XIA Kai-zhe, WU Ai-ling.
A Hybrid TCN-Transformer Architecture for Braking Modeling of Medium-low Speed Maglev Trains[J].
Science Technology and Engineering, 2026, 26(6): 2579-2591 DOI:10.12404/j.issn.1671-1815.2503057
ZhouYanli, WuYue, LuRongxiu, et al. Intelligent compensation modeling method for operation process of medium and low-speed maglev trains[J]. Journal of the China Railway Society, 2025, 47(3): 118-126.
ZhaoYiyun, LinHui, LiBingqiang, et al. Optimization control method for clamping force of electromechanical brake system of tram[J]. Electric Machines and Control, 2024, 28(7): 141-151.
ZhaoYiyun, LinHui, LiBingqiang, et al. Clamping force sensorless control strategies for electromechanical brake systems[J]. Journal of Beijing University of Aeronautics and Astronautics, 2023, 49(10): 2711-2720.
LiMaoqing, WangYiqing, GaoYunbo, et al. Adaptive control of electric multiple unit foraccurate stopping based on multi-mass model[J]. Control Theory & Applications, 2022, 39(8): 1579-1586.
LiMaoqing, ZhuShengjie, LinJunting, et al. Robust control for accurate stopping of urban rail vehicles with unknown time-varying parameters[J]. Journal of the China Railway Society, 2024, 46(7): 65-72.
WuXiaochun, JinZeling. Research on train energy saving control strategy based on DDPG algorithm[J]. Journal of Railway Science and Engineering, 2023, 20(2): 483-493.
LüWeixi, MinYongzhi, WangGuo, et al. Trajectory optimization method of freight suspended monorail train in complex environment[J]. Journal of Shanghai JiaoTong University, 2025.DOI:10.16183/j.cnki.jsjtu.2024.412.
[15]
MienyeI D, SwartT G, and ObaidoG. Recurrent neural networks: a comprehensive review of architectures, variants, and applications[J]. Information, 2024, 15(9): 1-34.
[16]
FournierQ, CaronG M, and AloiseD. Recurrent neural networks: a practical survey on faster and lighter transformers[J]. Acm Computing Surveys, 2024, 55: 1-40.
[17]
LiD F, LinK X, LiX T, et al. Improved sales time series predictions using deep neural networks with spatiotemporal dynamic pattern acquisition mechanism[J]. Information Processing & Management, 2022, 59(4). DOI: 10.1016/j.ipm.2022.102987.
[18]
VaswaniA, ShazeerN, and ParmarN, et al. Attention is all you need[J]. Advances in Neural Information Processing Systems, 2017, 30. DOI: 10.48550/arXiv.1706.03762.
ZhouFeng, WangShipu, ZhangKunpeng. A short-term traffic flow prediction model for urban road network based on graph transformer network[J]. Science Technology and Engineering, 2024, 24(10): 4307-4316.
XuRuiqi, LiuDandan. Non-invasive load monitoring based on multi-scale feature fusion and multi-head self-attention mechanism[J]. Science Technology and Engineering, 2024, 24(6): 2385-2395.
[23]
ZhaoW, GaoY, JiT, et al. Deep temporal convolutional networks for short-term traffic flow forecasting[J]. Expert Systems with Applications, 2019, 7: 114496-114507.
[24]
ZhaoW, GaoY, JiT, et al. A comprehensive survey on applications of transformers for deep learning tasks[J]. IEEE Access, 2023, 241: 122666.
[25]
MaksimsK, SamadD M. Deep imputation of missing values in time series health data: a review with benchmarking[J]. Journal of Biomedical Informatics, 2023, 144: 104440-104440.
[26]
BolikulovF, NasimovR, RashidovA, et al. Effective methods of categorical data encoding for artificial intelligence algorithms[J]. Mathematics, 2024, 12(16): 2553-2553.
ZhangYoubing, WangJianmin, ZhangGuozhen, et al. Accuracy and efficiency analysis of braking curve calculation for high-speed railway train[J]. Journal of Railway Science and Engineering, 2022, 19(1): 10-18.
[31]
ZhaoW T, DingJ M, ZhangQ S, et al. A distributed multiparticle precise stopping control model based on the distributed model predictive control algorithm for high-speed trains[J]. Journal of Dynamic Systems, Measurement, and Control, 2023, 145(11): DOI: 10.1115/1.4063015.