| Dual-Layer Predictive Energy Control in High-Speed Trains Using Adaptive Observers |
| Vo Thanh Ha, Bao Dan |
| Faculty of Electrical and Electronic Engineering, University of Transport and Communications, Hanoi, 100000, Vietnam |
|
|
|
Received: June 27, 2025; Revised: September 1, 2025 Accepted: September 8, 2025. Published online: September 30, 2025. |
|
|
|
| ABSTRACT |
|
Energy efficiency is crucial in modern high-speed rail, particularly with the emergence of hybrid energy storage for sustainable transportation. Traditional energy management methods, such as rule-based control and heuristics, lack scalability and adaptability for real-time, multi-unit operations. Although model predictive control (MPC) effectively manages energy flow within constraints, most implementations rely on centralised, single-layer architectures that struggle with flexibility and practical uncertainties. This paper presents a hierarchical energy management framework featuring a two-layer MPC structure and an adaptive state observer. The upper layer optimises long-term energy distribution, while the lower layer manages real-time torque tracking and power control. The adaptive observer enhances model accuracy and control reliability by estimating unmeasurable internal states, such as battery resistance and thermal degradation. Simulations using a high-fidelity train model demonstrate that the proposed framework enhances energy efficiency, mitigates battery stress, and maintains operational stability under varying loads. Its computational efficiency supports real-time onboard use, providing a scalable and robust solution for next-generation intelligent rail systems. |
| Key Words:
High-speed train · Energy management system · Hybrid energy storage · Model predictive control (MPC) · Hierarchical control · Adaptive observer |
|
Corresponding Author.
Vo Thanh Ha , Email. 
Preview 
Full text via DOI 
Download Citation
Print
