Bio-inspired Smart Attachment and Adaptive Neuromechanical Control for Highly Efficient Locomotion and Adaptation to an Autonomous Climbing Robot

点击翻译按钮获取中文摘要

Climbing robots are able to move on smooth or rough ground, walls or ceilings, and have wide application prospects in public and national security areas, narrow space detection, and cityscape services. Legged climbing robots, having a strong adaptability to the environment, can cross the complex terrain surface and have good movement flexibility. At present, their control and movement generation mainly rely on engineering approaches where online learning and adaptation have not been fully integrated. This limits the overall performance of the robot to deal with a changing environment. In contrast, geckos, insects, and spiders can efficiently walk and climb on different substrate slope and roughness. They can be stably attached and crawl on unsteady vertical (such as walls) or inverted (such as ceilings) surfaces in the natural environment, relying primarily on mechanical claw interlocking and foot pad adhesion as well as on the synergistic effect of both. Furthermore, they can even quickly adapt their locomotion to a change of their environment or an unknown situation. The versatile and adaptive abilities are the result of an integration of several ingredients embedded in their sensorimotor loop. Biological studies reveal that the key ingredients include biomechanics (e.g., adhesive pads and claws), sensory feedback, and adaptive neural control. Taking these ingredients into account for robot development together with in-depth studying and understanding attachment mechanisms of climbing animals in steady and unsteady states will lead to an adaptive, versatile, and efficient bio-inspired climbing robot. According to this, the purpose of the project is to 1) study and analyze the adhesive mechanisms of climbing animals based on the synergic action of claw mechanical interlocking and pad adhesion, and the multiple-peeling strategy of climbing animals for smart attachment control, 2) perform biological experiments, investigate locomotion behavior and its reaction forces of sprawled posture animals by using the former developed animal full-space motion behavior and reaction force testing system, and establish dynamic model of quadruped, 3) develop adaptive neuromechanical control with sensory feedback and online adaptation for stable and efficient climbing strategies, and 4) develop a bionic foot with adhesive pads and active claws. The developed bionic foot and adaptive neuromechanical control will be implemented on our autonomous climbing robot to achieve highly efficient locomotion and adaptation on changing incline angles and walls with smooth and different roughness substrates.