LPV-MPC for Lateral Control in Full-Scale Autonomous Racing
Hassan Jardali, Ihab S. Mohamed, Durgakant Pushp, Lantao Liu
AI summary
Problem
High-speed autonomous racing demands precise lateral control that balances nonlinear vehicle dynamics, actuator limits, and environmental disturbances within strict computational and track-time constraints. Existing controllers often sacrifice real-world validation, computational efficiency, or explicit constraint handling.
Approach
The authors design a Linear Parameter-Varying Model Predictive Controller that dynamically adjusts its internal model based on real-time longitudinal speed, track curvature, and road banking. This scheduling allows the controller to explicitly handle steering constraints and actuator dynamics while minimizing tracking errors.
Key results
- LPV-MPC framework scheduled by speed, curvature, and banking
- Stable high-speed validation exceeding 160 mph on a full-scale AV
- Iterative EM-based system identification for Pacejka tire parameters
- Open-source Python implementation released for the research community
Why it matters
Delivers a computationally efficient, constraint-aware control strategy validated at extreme speeds, providing a practical benchmark and open framework for high-performance autonomous vehicle development.
Abstract
Autonomous racing has attracted significant attention recently, presenting challenges in selecting an optimal controller that operates within the onboard system’s computational limits and meets operational constraints such as limited track time and high costs. This paper introduces a Linear Parameter- Varying Model Predictive Controller (LPV-MPC) for lateral control. Implemented on an IAC AV-24, the controller achieved stable performance at speeds exceeding 160 mph (71.5 m s−1). We detail the controller design, the methodology for extracting model parameters, and key system-level and implementation considerations. Additionally, we report results from our final race run, providing a comprehensive analysis of both vehicle dynamics and controller performance. Moreover, a Python implementation of the framework can be accessed here: https: //tinyurl.com/LPV-MPC-acados