Diffusion-Residual Model Predictive Steering Control for Vehicle Stabilization at the Limit of Handling under Model Uncertainty
2026-07-11 • Robotics
Robotics
AI summaryⓘ
The authors address the challenge of controlling a car's yaw (turning behavior) safely under changing conditions. They create a model that learns uncertainties in how the car should turn, based on the steering input, and uses this to adjust both the target yaw and safety limits ahead of time. This method helps the car anticipate problems instead of reacting too late, improving stability especially on slippery roads. Their approach runs efficiently on standard hardware and improves handling compared to traditional methods that use fixed assumptions.
Model Predictive Control (MPC)Yaw rateStable-handling envelopeConditional diffusion modelResidual modelingChance constraintsVehicle dynamicsSide-slipDirectional stabilityReal-time control
Authors
Bongsob Song
Abstract
At the limit of handling, a stabilizing MPC depends on the yaw-rate reference it tracks and the stable-handling envelope it enforces, both operating-point-dependent and unknown a priori, so fixed or worst-case settings are either too conservative or unsafe. We learn this uncertainty with a conditional diffusion residual model and apply it to the controller's reference and constraints rather than its control law. Conditioned on the steering command, the model returns the residual's mean and a predictive spread: the mean re-sizes the tracked yaw reference, while the spread, propagated over the prediction horizon, tightens the stable-handling envelope through a one-sided chance back-off. Together these form the proposed diffusion-residual MPC (D-res), so caution is anticipated ahead of the tracking error rather than corrected after it by a high-gain loop. Because only two moments per command are needed, the generator is tabulated offline and the online controller adds a single table lookup to the baseline MPC, with no in-loop diffusion; it runs within the 100 Hz budget on an NVIDIA Jetson AGX Xavier (worst-case 4.08 ms per step). Across a 7-DOF model and high-fidelity CarMaker co-simulation spanning vehicle, tire, road, and maneuver diversity, D-res reduces peak side-slip where the fixed bicycle model is least accurate and restores directional stability on low-friction maneuvers, where the fixed reference over-commands the available grip.