Answers are grounded in this paper and cite their sources.
With the rapid expansion of high-speed and heavy-haul railways in China, traditional railway stabilization operations—an open-loop approach heavily dependent on worker experience—are increasingly unable to meet the growing demands of railway maintenance in terms of both efficiency and quality. This paper presents the working principle of railway stabilization in detail and proposes, for the first time, a nonlinear model predictive control (NMPC) strategy based on spatial-domain receding-horizon optimization. Specifically, using the MFBD and DEM theories together with the Box-Behnken experimental design method, the evolution laws of the mechanical properties of an initially stable ballast bed under variable stabilization conditions are simulated. A nonlinear time-varying model of the lateral resistance of the ballast bed is established by a hybrid modeling approach combining polynomial response surface methodology (RSM) and a deep neural network. Subsequently, in view of the spatially moving characteristics of railway stabilization operations, an NMPC architecture integrating queue state management, constraint prioritization, and spatial-domain receding-horizon optimization is proposed. Finally, simulation and experimental results show that after NMPC operation, the average lateral resistance of the ballast bed reaches 14 kN, the one-pass compliance rate exceeds 90%, the operation time is reduced by 40%, and the improvement rate of lateral resistance is 2.4 times that of traditional operations. Compared with traditional operation methods, the proposed control strategy achieves better improvement in the quality state of the ballast bed and higher operational efficiency. The findings of this study can provide guidance for scientific stabilization operations and promote the theoretical development of railway maintenance.
No takes yet. Share an insight, caveat, or question.
Chen et al. (2026) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: