Short Inter-Robot Distance Following Based on Collision-Free Nonlinear Model Predictive Control

Self-reconfigurable robots aim to develop intelli-gent robotic systems with shape variability and task adaptability. This technology employs modular design, allowing robots to autonomously adjust their shape and structure according to specific tasks. For multiple self-reconfigurable robots using mecanum wheels as their mobile platform, a collision-free nonlinear model predictive control (NMPC) method is adopted to provide control strategies, enabling them to accomplish short-distance tracking tasks in real-time environments. An external radar module is utilized for global robot localization and obtaining high-precision pose information. Experiments conducted in a 3D simulation environment validate the effectiveness of this approach. In practical applications, overcoming various challenges such as hardware design, communication, and perception is essential to achieve robustness and reliability in self-reconfigurable robot systems.