TY - BOOK ID - 48211729 TI - Reinforcement learning of bimanual robot skills AU - Colomé, Adrià. AU - Torras, Carme. PY - 2020 SN - 3030263266 3030263258 PB - Cham : Springer International Publishing : Imprint: Springer, DB - UniCat KW - Artificial intelligence. KW - Robotics and Automation. KW - Artificial Intelligence. KW - Control and Systems Theory. KW - AI (Artificial intelligence) KW - Artificial thinking KW - Electronic brains KW - Intellectronics KW - Intelligence, Artificial KW - Intelligent machines KW - Machine intelligence KW - Thinking, Artificial KW - Bionics KW - Cognitive science KW - Digital computer simulation KW - Electronic data processing KW - Logic machines KW - Machine theory KW - Self-organizing systems KW - Simulation methods KW - Fifth generation computers KW - Neural computers KW - Robots KW - Kinematics. KW - Kinematics of robots KW - Robot kinematics KW - Machinery, Kinematics of KW - Robotics. KW - Automation. KW - Control engineering. KW - Control engineering KW - Control equipment KW - Control theory KW - Engineering instruments KW - Automation KW - Programmable controllers KW - Automatic factories KW - Automatic production KW - Computer control KW - Engineering cybernetics KW - Factories KW - Industrial engineering KW - Mechanization KW - Assembly-line methods KW - Automatic control KW - Automatic machinery KW - CAD/CAM systems KW - Robotics KW - Machine learning. KW - Dynamics. KW - Dynamics KW - Learning, Machine KW - Artificial intelligence UR - https://www.unicat.be/uniCat?func=search&query=sysid:48211729 AB - This book tackles all the stages and mechanisms involved in the learning of manipulation tasks by bimanual robots in unstructured settings, as it can be the task of folding clothes. The first part describes how to build an integrated system, capable of properly handling the kinematics and dynamics of the robot along the learning process. It proposes practical enhancements to closed-loop inverse kinematics for redundant robots, a procedure to position the two arms to maximize workspace manipulability, and a dynamic model together with a disturbance observer to achieve compliant control and safe robot behavior. In the second part, methods for robot motion learning based on movement primitives and direct policy search algorithms are presented. To improve sampling efficiency and accelerate learning without deteriorating solution quality, techniques for dimensionality reduction, for exploiting low-performing samples, and for contextualization and adaptability to changing situations are proposed. In sum, the reader will find in this comprehensive exposition the relevant knowledge in different areas required to build a complete framework for model-free, compliant, coordinated robot motion learning. ER -