EUVIMEDEUVIMED entdecken
Crossref

Sleeper-Encoded Localization Method: A Mileage Precision Localization Approach for Railway Maintenance

Zhiheng Wang, Shujie Sun, Xuanrui Zhang, Shengchun Wang, Liang Wang

IEEE Sensors Journal · 2026-08 · 26 (16) · 24578-24586

DOI: 10.1109/jsen.2026.3709488 ↗

Kurz gefasst

Fachartikel in der Zeitschrift IEEE Sensors Journal, 2026-08. Verfasst von Zhiheng Wang und 4 weiteren Autor:innen. Die Arbeit nennt 23 Literaturangaben. Gefördert von: National Natural Science Foundation of China, Key Research Project of China Academy of Railway Sciences Corporation Ltd, Science and Technology Research and Development Plan of China State Railway Group Company Ltd..

AbstractBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

High-precision mileage localization is a critical prerequisite for railway maintenance, requiring inspection data to be mapped to physical targets with high deterministic precision. However, bridging the gap between continuous sensor measurements and discrete track infrastructure remains challenging. Conventional odometry-based methods suffer from cumulative drift, while state-of-the-art deep learning (DL) pipelines, despite fast inference, often face system-level latency dominated by CPU-bound data preparation (e.g., voxelization or projection). To address these issues, this article proposes sleeper-encoded localization (SEL), a bounded-drift localization paradigm that reformulates along-track positioning as a hierarchical geometric filtering process. A micro-scale geometric filter segments sleepers from decoupled 2-D profiles (with theoretical validity on curved tracks), while a macroscale topological filter enforces global indexing consistency using physically plausible spacing constraints. Experiments on an operational high-speed railway segment show that SEL achieves 0.22 ms/profile on a standard CPU, effectively mitigating the data restructuring bottleneck commonly observed in 3-D DL pipelines. On the evaluated trajectory, SEL maintains consistent sleeper indexing and constrains the longitudinal error within a bounded interval determined by sleeper spacing and local detection ambiguity. In addition, stress tests with physically motivated degradations (e.g., sparsity and localized occlusion) illustrate robustness in realistic regimes and graceful degradation as conditions become more challenging. Overall, SEL provides a fast and lightweight sleeper-indexed longitudinal reference that complements odometer/GNSS/INS-based absolute positioning in resource-constrained maintenance workflows, rather than serving as an independent absolute mileage measurement sensor.

Förderung

  • National Natural Science Foundation of China
  • Key Research Project of China Academy of Railway Sciences Corporation Ltd
  • Science and Technology Research and Development Plan of China State Railway Group Company Ltd.

Zitieren

APA

Wang, Z., Sun, S., Zhang, X., Wang, S., & Wang, L. (2026). Sleeper-Encoded Localization Method: A Mileage Precision Localization Approach for Railway Maintenance. IEEE Sensors Journal, 26(16), 24578-24586. https://doi.org/10.1109/jsen.2026.3709488

Vancouver

Wang Z, Sun S, Zhang X, Wang S, Wang L. Sleeper-Encoded Localization Method: A Mileage Precision Localization Approach for Railway Maintenance. IEEE Sensors Journal. 2026;26(16):24578-24586. doi:10.1109/jsen.2026.3709488

BibTeX
@article{Wang2026,
  author = {Wang, Zhiheng and Sun, Shujie and Zhang, Xuanrui and Wang, Shengchun and Wang, Liang},
  title = {Sleeper-Encoded Localization Method: A Mileage Precision Localization Approach for Railway Maintenance},
  journal = {IEEE Sensors Journal},
  year = {2026},
  volume = {26},
  number = {16},
  pages = {24578--24586},
  doi = {10.1109/jsen.2026.3709488}
}
Kostenlos starten →