Related publications
- AD-MPCC: Adaptive Differentiable Model Predictive Contouring Control for Autonomous Racing2026 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS), 2026
Overview
Autonomous racing controllers must respond to changing road conditions while balancing progress, tracking, and safety. AD-MPCC combines a differentiable model predictive contouring controller with online estimation of tire-road interaction parameters.
Method
A regularized moving-horizon estimator updates a parameterized Pacejka tire model from recent observations. Differentiating through the controller enables the learning of objective weights against longer-horizon performance goals; a Pacejka-informed model then adapts those weights online.
Evaluation
The paper evaluates single-surface and mixed-surface racing in simulation, comparing safety and lap time with baseline controllers.
Contributors
Nam T. Nguyen, Binh Nguyen, Ahmad Amine, Thanh Vo-Duy, Rahul Mangharam, and Truong X. Nghiem.