Autonomous Drifting with 3 Minutes of Data via Learned Tire Models

TL;DR

This paper introduces a neural ODE-based tire force model that, with less than three minutes of data, significantly improves autonomous drifting performance and control smoothness in a real vehicle.

Contribution

The paper presents a novel neural ODE tire model that captures complex tire dynamics using minimal data, integrated into a control framework for high-performance autonomous drifting.

Findings

Achieved high-performance drifting with less than three minutes of data

Outperformed benchmark models with 4x better tracking accuracy

Demonstrated smoother control inputs and faster computation times

Abstract

Near the limits of adhesion, the forces generated by a tire are nonlinear and intricately coupled. Efficient and accurate modelling in this region could improve safety, especially in emergency situations where high forces are required. To this end, we propose a novel family of tire force models based on neural ordinary differential equations and a neural-ExpTanh parameterization. These models are designed to satisfy physically insightful assumptions while also having sufficient fidelity to capture higher-order effects directly from vehicle state measurements. They are used as drop-in replacements for an analytical brush tire model in an existing nonlinear model predictive control framework. Experiments with a customized Toyota Supra show that scarce amounts of driving data -- less than three minutes -- is sufficient to achieve high-performance autonomous drifting on various trajectories with speeds up to 45mph. Comparisons with the benchmark model show a $4 \times$ improvement in tracking performance, smoother control inputs, and faster and more consistent computation time.