Robot Learning Human Skills and Intelligent Control Design

ISBN-13: 9780367634360 / Angielski / Twarda / 2021 / 174 str.

Chenguang Yang; Chao Zeng; Jianwei Zhang

Opis

This book focusses on robotic skill learning and intelligent control for robotic manipulators including enabling of robots to efficiently learn motor and stiffness/force regulation policies from humans. It explains transfer of human limb impedance control strategies to the robots so that the adaptive impedance control for the robot can be realized.

Spis treści

Chapter 1 Introduction1.1Overview of sEMG-based stiffness transfer1.2Overview of robot learning motion skills from humans1.3Overview of robot intelligent control designChapter 2 Robot platforms and software systems2.1Baxter robot2.2Nao robot2.3KUKA LBR iiwa robot2.4Kinect camera2.5MYO Armband2.6Leap Motion2.7Oculus Rift DK 22.8MATLAB Robotics Toolbox2.9CoppeliaSim2.10GazeboChapter 3 Human-robot stiffness transfer based on sEMG signals3.1Introduction3.2Brief introduction of sEMG signals3.3Calculation of human arm Jacobian matrix3.4Stiffness estimation3.5Interface design for stiffness transfer3.6Human-robot stiffness mapping3.7Stiffness transfer for various tasks3.8Conclusion Chapter 4 Learning and Generalisation of Variable Impedance Skills4.1Introduction4.2Overview of the framework4.3Trajectory segmentation4.4Trajectory alignment methods4.5Dynamical movement primitives4.6Modeling of impedance skills4.7Experimental study4.8ConclusionChapter 5 Learning human skills from multimodal demonstration5.1Introduction5.2System Describtion5.3HSMM-GMR Model Description5.4Impedance Controller in Task Space5.5Experimental Study5.6Conclusion Chapter 6 Skill Modeling based on Extreme Learning Machine6.1Introduction6.2System of teleoperation-based robotic learning6.3Human/robot joint angle calculation using Kinect camera6.4Processing of demonstration data6.5Skill modeling using extreme learning machine6.6Experimental study6.7Conclusion Chapter 7 Neural Network Enhanced Robot Manipulator Control7.1Introduction7.2Problem description7.3Learning from multiple demonstrations7.4Neural networks techniques7.5Robot manipulator controller design7.6Experimental study7.7ConclusionReferences