EUVIMEDEUVIMED entdecken
Open Access · CC BY-NC-ND 4.0

Reinforcement Learning-Based Adaptive Gain Tuning of Terminal Super-Twisting SMC for Lane-Change Control in Autonomous Vehicles

Mohammadreza Maleki, Forough Sadat Razavi, Amir Taghavipour

IEEE Access · 2025 · 13 · 197206-197218

DOI: 10.1109/access.2025.3634379 ↗

Kurz gefasst

Fachartikel in der Zeitschrift IEEE Access, 2025. Verfasst von Mohammadreza Maleki und 2 weiteren Autor:innen. Die Arbeit wurde bisher 7-mal zitiert (Crossref) und nennt 39 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY-NC-ND 4.0).

AbstractBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

This paper presents a reinforcement learning–based adaptive gain tuning strategy for a Terminal Super-Twisting Sliding Mode Controller (TST-SMC) applied to autonomous vehicle lane-change maneuvers. The controller gains are adapted per-maneuver using a Twin-Delayed Deep Deterministic Policy Gradient (TD3) algorithm, conditioned on vehicle velocity. The vehicle model, including tire–road interactions, is developed in MATLAB/Simulink-python. Simulation results show that the proposed method reduces lateral position RMSE by up to 28%, yaw RMSE by 20%, and maximum lateral deviation by 33% compared with fixed-gain TST-SMC, while lowering steering effort by about 6-10% across the tested velocity range. These improvements demonstrate that reinforcement learning–based adaptation enhances the robustness and efficiency of sliding mode control under diverse operating conditions.

Zitieren

APA

Maleki, M., Razavi, F. S., & Taghavipour, A. (2025). Reinforcement Learning-Based Adaptive Gain Tuning of Terminal Super-Twisting SMC for Lane-Change Control in Autonomous Vehicles. IEEE Access, 13, 197206-197218. https://doi.org/10.1109/access.2025.3634379

Vancouver

Maleki M, Razavi FS, Taghavipour A. Reinforcement Learning-Based Adaptive Gain Tuning of Terminal Super-Twisting SMC for Lane-Change Control in Autonomous Vehicles. IEEE Access. 2025;13:197206-197218. doi:10.1109/access.2025.3634379

BibTeX
@article{Maleki2025,
  author = {Maleki, Mohammadreza and Razavi, Forough Sadat and Taghavipour, Amir},
  title = {Reinforcement Learning-Based Adaptive Gain Tuning of Terminal Super-Twisting SMC for Lane-Change Control in Autonomous Vehicles},
  journal = {IEEE Access},
  year = {2025},
  volume = {13},
  pages = {197206--197218},
  doi = {10.1109/access.2025.3634379}
}
Kostenlos starten →