Safe and Robust Motion Planning for Autonomous Navigation of Quadruped Robots in Cluttered Environments

Quadruped robots, with their superior terrain adaptability and flexible movement capabilities, demonstrate greater application potential in complex environments compared to traditional ground robots. However, their non-negligible body shape and anisotropic motion characteristics complicate the achievement of high-precision motion planning and autonomous navigation. In this paper, we propose a safe and robust motion planning system tailored for autonomous navigation of quadruped robots in cluttered environments. We adopt a hierarchical architecture and decompose the planning process into front-end searching and back-end optimization. In the front-end searching stage, the robot finds a smooth, feasible, and energy-efficient initial trajectory with safety consideration. In the back-end optimization stage, we leverage B-splines to enhance the trajectory smoothness, safety, and motion stability. Finally, the time allocation is fine-tuned through iterative refinement, ensuring the feasibility of the optimized trajectory. Our method is extensively validated in challenging simulations as well as in real-world testing environments, benchmark comparisons also demonstrate the improved performance of our method.