Motion-Based Slide Correction Controller for Autonomous Vehicles

The automotive industry has successfully deployed robust sliding stability control systems. However, the rapidly expanding and heterogeneous field of wheeled robotics lacks an equivalent, standardised solution for managing slides. Existing methods are often model-dependent, requiring extensive, platform-specific characterisation that is impractical at the scale of robotic development. This thesis addresses this gap by developing a universal, motion-based Slide Correction Controller (SCC) inspired by the intuitive, model-free techniques of skilled human drivers. The core philosophy is to correct slides by focusing on the vehicle’s observable kinematics rather than its underlying dynamics. The foundation of the SCC is a universal mathematical framework that defines a slide as the error between the vehicle’s actual and expected body motion, calculated from wheel kinematics (eslide = vactual − vmodelled). This platform-agnostic principle enables the detection of all primary slide types: wheel-lock, wheel-spin, oversteer, and understeer. The SCC architecture integrates two specialised modules: a Longitudinal Sliding Mode Controller (LSMC) that manages translational slip using non-linear control, and a Yaw Rate Controller (YRC) that addresses rotational slip through a novel, model-free implementa-tion of active counter-steering. High-fidelity simulation environments for two kinematically distinct platforms, a lightweight Ackermann-steered vehicle (RoboRacer) and a heavy differential-drive robot (Voyager), were developed. The SCC was created using these en-vironments, with systematic parameter tuning and robustness testing against variations in friction, mass, and vehicle geometry. The controller’s platform-agnostic design was tested by deploying the SCC, with minimal retuning. The YRC’s performance was validated on both real-world hardware platforms, where it successfully corrected slides induced by aggressive manoeuvres and external disturbances, confirming its cause-agnostic capabilities. The results demonstrate that the integrated SCC is essential for navigating complex, compound slide events (such as emergency avoidance manoeuvres or sudden friction transitions) where individual longitudinal or lateral controllers fail. This research concludes that the motion-based control philosophy provides an effective foundation for a universal slide correction system, offering a robust safety layer that can enhance the stability of a wide range of autonomous wheeled vehicles.

Grantor dc:publisher

Stellenbosch : Stellenbosch University

Year dc:date.issued

2026

Author dc:creator

• Chetty, Brandon

Advisors dc:contributor.advisor

• Jordaan, H. W.
• Evans, B. D.

Language dc:language.iso

en

Repository record dc:identifier.uri

https://scholar.sun.ac.za/handle/10019.1/135679

OAI identifier oai:identifier

oai:scholar.sun.ac.za:10019.1/135679