Fixed-time convergent lateral control of autonomous vehicles based on geometric controllers

Abstract Vehicle dynamics modelling plays a crucial role in developing accurate and robust controllers for autonomous vehicles (AVs). The applied vehicle models range from simple geometric ones to sophisticated nonlinear models involving tyre behaviour. In this study, a nonlinear controller is designed by combining geometric models of pure pursuit (PP) and Stanley (ST) controllers. This combination benefits from the preview information of the PP controller and the cross-track error information of the ST controller. Based on this feature, a fixed-time convergent controller is proposed for the lateral control of an AV. A novel tuning algorithm is designed to effectively weight the contribution of lateral cross-track error and preview heading error to overall performance of the controller. This tuning algorithm enhances the driving speed of geometric controllers to over 100  $$\text{km}/\text{h}$$ km / h and considerably reduce the error in high-speed emergency manoeuvres. Simulations are carried out and comparisons are made with several algorithms. The results outline superior performance of the proposed controller over geometric and dynamic model-based algorithms. Last but not least, the proposed algorithm paves the way for introducing capable nonlinear controllers using the geometric vehicle models.