Contribution

Simulation-Driven Optimization of Stanley Controller Gains for Enhanced Tracking in Autonomous Navigation Robots

---

Abstract

Fine-tuning controllers for robotic systems is a tedious process that often requires significant time for convergence and can lead to mechanical component wear. Having an accurate simulation of the robotic system and its Environment can help reduce this effort and accelerate the tuning process. This work presents an optimization-based approach that leverages simulations to optimize control parameters before transferring them to a real mobile robot, significantly reducing fine-tuning effort and the need for extensive real-world testing. The method follows a two-stage process: first, calibrating the simulator to closely replicate the mobile robot’s trajectory, and second, using the refined simulation to optimize the Stanley controller’s gains. By aligning the simulator’s behavior with real-world performance, we ensure that control tuning is both effective and time-efficient, allowing optimized parameters to be directly applied to the real system. The methodology is validated through experiments comparing simulated and real-world trajectories, demonstrating that the optimized gains improve tracking accuracy. Additionally, we provide an estimation of the achieved improvements, including tracking error reduction, time savings, and energy consumption minimized by our approach, highlighting its efficiency in the fine-tuning process.

Keywords: Autonomous navigation, Stanley controller, simulator optimization, control tuning, simulation-to-reality transfer, parameter optimization

How to Cite:

Pérez-Villeda, H., Mühlbacher, C. & Mautner-Lassnig, K., (2025) “Simulation-Driven Optimization of Stanley Controller Gains for Enhanced Tracking in Autonomous Navigation Robots”, ARW Proceedings 25(1), 85-90. doi: https://doi.org/10.34749/3061-0710.2025.14