Whole-body trajectory optimization for skid steer wheeled-legged vehicles

Wheel-legged vehicles combine the advantage of high-speed capability of wheeled locomotion with the superior mobility of legged mechanisms, demonstrating exceptional performance in structured environments. However, these hybrid systems face several challenges regarding energy efficiency and motion control complexity when operating in unstructured terrains. To address these challenges, this study proposes a real-time whole-body trajectory optimization framework, designed for wheel-legged hybrid vehicles, with optimized performance in challenging environments, such as staircases, slopes, and construction sites. The proposed framework innovatively integrates the obstacle negotiation capability of legged robots with the slip-steering advantage of wheeled platforms, providing distinctive benefits for various applications, including search-and-rescue missions and planetary exploration. Unlike conventional wheel-legged designs, the proposed LegoWheel platform incorporates a streamlined mechanical architecture that eliminates redundant hip abduction-adduction (HAA) joints while realizing steering through differential wheel speed control. The unified optimization framework simultaneously considers wheeled and legged dynamics, providing optimal trajectories for joints and wheels, while addressing body motion, contact forces, and wheel torque distribution. The proposed framework is verified by extensive experiments across various challenging terrains, including staircases and trenches. The results demonstrate that the proposed LegoWheel can achieve significant improvements in energy efficiency for specific operational scenarios compared to traditional wheel-legged robots, while maintaining the robust obstacle-crossing feature and high-speed stability. The key innovations of this study include: (1) an innovative minimalist wheel-legged architecture that reduces mechanical complexity; (2) a unified optimization framework for hybrid locomotion dynamics; and (3) experimental validation of energy-efficient navigation in real-world unstructured environments. These advancements collectively enhance operational efficiency and performance in demanding conditions.