Abstract:Autonomous vehicles that can reliably operate outside the stable handling limits would have access to a wider range of maneuvers in emergencies, improving overall safety. To that end, this paper presents a novel Nonlinear MPC approach for vehicle control with deeply saturated rear tires. Longitudinal slip management is elevated from the chassis control layer into the optimisation problem by using a coupled-slip tire model, and explicitly including wheelspeed dynamics. Terminal costs on sideslip stability help compensate for the finite horizon, while road bounds and static obstacles are encoded using slack constraints. Experiments on a racetrack with a modified Toyota GR Supra validate the controller's ability to smoothly transition from dynamic, non-equilibrium drifting to grip driving. Further experiments demonstrate robustness to significant longitudinal force and wheelspeed disturbances, and showcase the controller flexibly transitioning in and out of the sliding tire regime to balance slack constraints with tracking objectives.

What problem does this paper attempt to address?

The core problem that this paper attempts to solve is: **How to enable autonomous vehicles to operate reliably beyond the stable handling limits in order to deal with emergency situations and thus improve overall safety**. Specifically, the paper proposes a new Non - Linear Model Predictive Control (NMPC) method for controlling vehicles during high - transient autonomous drifting, especially in the case of deep saturation of the rear - wheel tires. ### Background and Problem Description of the Paper Current autonomous vehicle controllers are usually limited by conservative open - loop stable conditions, which greatly limit the achievable dynamic states of the vehicle. However, experienced drivers often take advantage of the higher flexibility brought by fully saturated rear - wheel tires and skillfully balance multiple objectives during operation. If autonomous vehicles can operate in such extreme states like human drivers, they can access a wider range of maneuvers, thereby improving safety in many scenarios. ### Specific Problems 1. **Stability Challenges**: When the tires are fully slipping, the yaw rate and side - slip angle become open - loop unstable, which makes driving very difficult. 2. **Complex Dynamic Modeling**: It is necessary to consider the influence of vehicle configuration and road surface conditions on the minimum - time turn negotiation and stability. 3. **Multi - task Control**: Not only is it necessary to track the path, but also to keep the rotational state of the vehicle stable, which makes the problem an under - actuated problem (that is, the number of states is more than the number of inputs). 4. **Robustness in Reality**: The controller needs to show robustness in the presence of longitudinal force and wheel - speed perturbations. ### Solutions To solve these problems, the paper proposes the following innovations: - **Introduction of Non - Linear Model Predictive Control (NMPC)**: By using a high - fidelity vehicle model to capture complex dynamic behaviors while balancing the convergence time and the length of the prediction horizon. - **Tire Model Improvement**: Select an isotropic coupled - slip tire model, taking into account the coupling relationship between longitudinal and lateral slips, especially the behavior when the tires are fully slipping. - **Inclusion of Wheel - speed Dynamics in the Optimization Problem**: Incorporate the wheel - speed of the driven rear axle as an explicit vehicle state into the NMPC problem formulation, so that the controller can compensate for the real - time measured wheel - speed changes. - **Relaxed Constraints to Handle Boundary Conditions**: In order to prevent infeasibility problems when approaching the boundary, use large - error - bound relaxed constraints with large quadratic penalties. ### Experimental Verification The experiment was carried out on a commercial racetrack using a modified Toyota GR Supra (code - named "Keisuke"). The experiment consists of three parts: 1. Initial grip driving 2. Handbrake destabilization 3. Transient drifting The NMPC controller is responsible for controlling the vehicle in these three stages, demonstrating its ability to smoothly transition from dynamic unbalanced drifting to stable grip driving. The experimental results show that the controller can flexibly switch inside and outside the slipping - tire area to balance relaxed constraints and tracking objectives, and shows good robustness in the face of external perturbations. In conclusion, this paper aims to improve the safety of dealing with emergency situations by enabling autonomous vehicles to operate reliably under extreme conditions through the introduction of an advanced NMPC control framework.

Beyond the stable handling limits: nonlinear model predictive control for highly transient autonomous drifting

Combined Fast Control of Drifting State and Trajectory Tracking for Autonomous Vehicles Based on MPC Controller

Learning-Based Hierarchical Model Predictive Control for Drift Vehicles

Learning-Based MPC Controller for Drift Control of Autonomous Vehicles

An MPC-based Controller Framework for Agile Maneuvering of Autonomous Vehicles.

Nonlinear Model Predictive Control for Enhanced Path Tracking and Autonomous Drifting through Direct Yaw Moment Control and Rear-Wheel-Steering

NMPC-based controller for vehicle longitudinal and lateral stability enhancement under extreme driving conditions

Model Predictive Control for Integrated Lateral Stability

Coordinated longitudinal and lateral vehicle stability control based on the combined-slip tire model in the MPC framework

A Nonlinear Model Predictive Control for Automated Drifting with a Standard Passenger Vehicle

Coordinated Longitudinal and Lateral Stability Improvement for Electric Vehicles Based on a Real-Time NMPC Strategy

A hierarchical adaptive nonlinear model predictive control approach for maximizing tire force usage in autonomous vehicles

Coupled Lateral and Longitudinal Control for Trajectory Tracking, Lateral Stability, and Rollover Prevention of High-Speed Automated Vehicles Using Minimum-Time Model Predictive Control

Risk-Averse Model Predictive Control for Racing in Adverse Conditions

Integrated Longitudinal and Lateral Vehicle Stability Control for Extreme Conditions With Safety Dynamic Requirements Analysis

Model predictive control for integrated longitudinal and lateral stability of electric vehicles with in‐wheel motors

A novel model predictive controller for the drifting vehicle to track a circular trajectory

Nonlinear MPC-based Slip Control for Electric Vehicles with Vehicle Safety Constraints

Efficient Nonlinear Model Predictive Control of Automated Vehicles

Integrated longitudinal and lateral control for autonomous vehicles with active load transfer strategy at the handling limits

Switched model predictive controller for path tracking of autonomous vehicle considering rollover stability