Physics-Based Robust, Adaptive, and Scalable Control Algorithms for Autonomous Ground Vehicles Operating at High-Speed in Adversarial Environments

High-speed maneuverability for off-road autonomous ground vehicles in unprepared, hostile environments continues to be a challenging problem, especially due to i) the lack of terramechanics models suitable for a high-speed dynamic response, ii) the need for real-time sensing of the environment and of the vehicle (or vehicle-terrain) parameters, and iii) the absence of fast trajectory planners and robust controllers that account for uncertainties in both the terrain’s models and vehicle’s model both in nominal and off-nominal conditions.

The goal of this project is to address these problems by advancing the state-of-the-art in four areas, namely terramechanics, ground vehicles parameter estimation, trajectory planning, and control theory, and smoothly integrating them. The theoretical solutions created for this work will be synthesized in innovative parameter estimation models, trajectory planning, and control algorithms that can be employed on multiple vehicles and multiple terrains while retaining satisfactory levels of performance. This research seeks to enable off-road maneuvers at high-speed, will guarantee higher levels of robustness to uncertainties on the ground-vehicle interactions and modeling errors in the vehicle’s actual dynamics, will guarantee adaptability to variable environmental conditions, and scalability to multiple vehicles.

Research objectives include:

1. Terramechanics and vehicle dynamics modeling for four-wheeled, four-passenger vehicles.
2. Parameter estimation. The main goal of this objective is to assess and provide in real-time the information needed for realistic, real-time trajectory planning (Objective 3) and robust adaptive control (Objective 4).
3. Real-time onboard trajectory planning in hostile environments.
4. Robust adaptive controllers for ground vehicles operating off-road.

Publications:

• White, H, Mukherjee, J., L’Afflitto, A., and Sandu, C. – “Terrain Sensing Techniques and Algorithms: A Review”, Paper no. 108, Proc. of The ISTVS 20th Int. and 9th Americas Conf. of the ISTVS, Sept. 27-29, 2021, Montreal, Canada (virtual).

#1.A91