Central pattern generators and hormone inspired messages: A hybrid control strategy to implement motor primitives on chain type modular reconfigurable robots

摘要

This paper presents a hybrid strategy for controlling a chain type modular robot to achieve a better interaction with the environment based on the combination of Central Pattern Generators (CPG) and Hormone inspired control. Locomotion movements of robot chains are generated by the local interaction of CPG running inside each module. Hormone messages are used to propagate sensory feedback information to the CPG thus providing a way to define general motor primitives regardless of the specific role of a module and the configuration of the robotic structure. An example set of motor primitives is defined using the hybrid approach and tested in a simulated environment. Our results showed that the integration of the motor primitives using the hybrid strategy allows the robot to travel through an obstacle-filled uneven terrain.