A Learning-based Planning and Control Framework for Inertia Drift Vehicles

TL;DR

This paper introduces a learning-based planning and control framework for inertia drift vehicles, using Bayesian optimization to improve transition smoothness and control accuracy during sharp corner maneuvers in autonomous racing.

Contribution

It presents a novel framework combining planning and control with Bayesian optimization to handle inertia drift transitions and model uncertainties.

Findings

Achieves smooth and stable inertia drift through sharp corners in simulation.

Enhances control robustness against modeling errors with learning-based policy.

Demonstrates improved path tracking and transition quality in simulation results.

Abstract

Inertia drift is a transitional maneuver between two sustained drift stages in opposite directions, which provides valuable insights for navigating consecutive sharp corners for autonomous racing.However, this can be a challenging scenario for the drift controller to handle rapid transitions between opposing sideslip angles while maintaining accurate path tracking. Moreover, accurate drift control depends on a high-fidelity vehicle model to derive drift equilibrium points and predict vehicle states, but this is often compromised by the strongly coupled longitudinal-lateral drift dynamics and unpredictable environmental variations. To address these challenges, this paper proposes a learning-based planning and control framework utilizing Bayesian optimization (BO), which develops a planning logic to ensure a smooth transition and minimal velocity loss between inertia and sustained drift phases. BO is further employed to learn a performance-driven control policy that mitigates modeling errors for enhanced system performance. Simulation results on an 8-shape reference path demonstrate that the proposed framework can achieve smooth and stable inertia drift through sharp corners.