Conclusions

Abstract

This book presents a hierarchical control architecture as steps toward developing limbless robots capable of 3D locomotion, fast reflex responses, as well as sophisticated reaction to environmental stimuli. The three main functionalities developed in the control system all serve important roles. First, the control system in limbless robots has the ability to generate 3D locomotive patterns so that the robots can move in environments. Second, the control system allows the limbless robots to fast respond to external stimuli via a reflex mechanism. And third, the control system provides the limbless robots with a mean for achieving deliberate goal-oriented behavior. Through simulations and experiments, the hierarchical control architecture is validated useful to ensure the limbless robots to intelligently and autonomously move in complex environments. As a highlight of the book, the hierarchical control architecture is designed to have two levels of control scheme. On the fine-scale level, a CPG controller is applied for gait generation and a reflex mechanism designed as a special extended pathway of the CPG model is used for fast response to external stimuli. On the large-scale level, learning algorithms are applied to study the mapping between the parameters of the CPG model and the sensory information gathered from the environment, so as to generate deliberate response behaviors to environmental changes. From the control point of view, the fine-scale level provides the large-scale level with the basic mobile ability, while the large-scale level in turn modifies the behavior generated by the fine-scale level by means of learned reaction rules.