Abstract
The effective and accurate walking is the biggest challenge for humanoid robot locomotion. In dynamic environments, such as RoboCup competition, not only must the speed of actions be high, but action switching must also be done almost immediately. This paper concentrates on two major behavior actions of humanoid soccer robot; straight walking and turning actions. Matsuoka central pattern generator model is used to generate trajectory actions. Both actions have their own parameters, which are obtained by comprehensive learning particle swarm optimization. By using these two actions and proper switching between them, robot can reach every point in the environment within a reasonable time. This paper tries to highlight the importance of action switching in movement maneuverability and proposes an effective solution for it. As transition from one action to another cannot be done in every posture of robot, the switching is done when robot is in double support phase. In this situation, membrane potential of each neuron has its minimum value and switching does not create big torque. The maximum time required for change action in this model is less than one step of robot. This method has been successfully implemented in rcssserver3d simulation server on NAO humanoid robot.
Sign-in/Register to access full text options 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a callMore From: International Journal of Machine Learning and Cybernetics
Disclaimer: All third-party content on this website/platform is and will remain the property of their respective owners and is provided on "as is" basis without any warranties, express or implied. Use of third-party content does not indicate any affiliation, sponsorship with or endorsement by them. Any references to third-party content is to identify the corresponding services and shall be considered fair use under The CopyrightLaw.
