Whole-Body Control for Multi-Contact Balancing of Humanoid Robots: Design and Experiments (Springer Tracts in Advanced Robotics, 143)

This book aims at providing algorithms for balance control of legged, torque-controlled humanoid robots. A humanoid robot normally uses the feet for locomotion. This paradigm is extended by addressing the challenge of multi-contact balancing, which allows a humanoid robot to exploit an arbitrary number of contacts for support. Using multiple contacts increases the size of the support polygon, which in turn leads to an increased robustness of the stance and to an increased kinematic workspace of the robot. Both are important features for facilitating a transition of humanoid robots from research laboratories to real-world applications, where they are confronted with multiple challenging scenarios, such as climbing stairs and ladders, traversing debris, handling heavy loads, or working in confined spaces. The distribution of forces and torques among the multiple contacts is a challenging aspect of the problem, which arises from the closed kinematic chain given by the robot and its environment. Series Editor’s Foreword Preface Contents Symbols and Abbreviations Abbreviations Sets Frames Matrices Scalars Vectors 1 Introduction 1.1 Motivation 1.2 Problem Description 1.3 State of the Art and Contributions 1.4 Publications and Outline References 2 Notation 3 Modeling 3.1 Rigid Body Transformations 3.1.1 Body Coordinates 3.1.2 World Coordinates 3.2 Robot Dynamics 3.2.1 Floating Base Dynamics 3.2.2 Centroidal Dynamics 3.3 Wrench Distribution Problem 3.4 Contact Model 3.4.1 The General Case of a Flat and Unilateral Contact 3.4.2 Approximation for Rectangular Contacts 3.4.3 Approximation for Point Contacts with Fixed Orientation 3.5 Overview of the Parameterizations of the Wrench Distribution Problem 3.6 Feasibility of the Overall Support Wrench References 4 Cartesian Compliance 4.1 General Description 4.2 Translational Stiffness 4.3 Rotational Stiffness 4.4 Damping References 5 Torque-Controlled Humanoid Robot TORO 5.1 Hardware Description 5.2 Base Frame Estimation References 6 Whole-Body Control for Multi-contact Balancing 6.1 Basic Concept (MCB) 6.1.1 Controller Derivation 6.1.2 Properties of the Controller 6.1.3 Experimental Evaluation 6.2 Null Space Controller 6.3 The Tracking Case: Following Dynamic Trajectories (MCB+) 6.3.1 Controller Derivation 6.3.2 Properties of the Controller 6.3.3 Experimental Evaluation 6.4 Passivity Control on Movable and Deformable Ground (MCB-PC) 6.4.1 Problem Description 6.4.2 Controller Derivation 6.4.3 Simplification of the Control Strategy for the Group Motion 6.4.4 Energy Tanks 6.4.5 Experimental Evaluation 6.5 Balancing While Performing High-Force Interaction Tasks (MCB-ACT) 6.5.1 Controller Derivation 6.5.2 Reacting to Contact Transitions 6.5.3 Experimental Evaluation 6.6 Interface for Providing Inputs from External Sources 6.7 Implementation 6.7.1 Increasing the Robustness of the Supporting Contacts 6.7.2 Formulation of the Constrained Quadratic Optimization 6.7.3 Procedure for Contact Transition References 7 Combining Multi-contact Balancing with Hierarchical Whole-Body Control 7.1 Multi-objective Control for Fixed Base Robots 7.2 Transfer to Robots with a Floating Base (HMCB) 7.2.1 Controller Derivation 7.2.2 Link to the MCB Control Approach for Multi-contact Balancing 7.2.3 Experimental Evaluation 7.3 Transfer to Confined Spaces 7.3.1 Theoretical Background 7.3.2 Experimental Evaluation References 8 Balance Control Based on Reduced Dynamic Models 8.1 Interaction Aware Balancing via LIPM 8.1.1 Linear Inverted Pendulum Model 8.1.2 Disturbance Observer 8.1.3 Controller Derivation 8.1.4 Position-Based ZMP Controller 8.1.5 Experimental Evaluation 8.2 Balancing Using a Locked Inertia Model 8.2.1 Controller Derivation 8.2.2 Evaluation in Simulation References 9 Applications 9.1 Combining Planning with Whole-Body Balancing 9.1.1 Autonomous Grasping 9.1.2 Stair Climbing 9.2 Combining Teleoperation with Whole-Body Balancing 9.2.1 Bimanual Teleoperation Using a Wearable, Ultralight Input Device 9.2.2 Teleoperation Using a Task-Relevant Haptic Interface 9.3 Real-World Applications: Employing Humanoid Robots in Aircraft Manufacturing References 10 Discussion and Conclusion References Appendix A Positive Definiteness of Matrix Potential Appendix B Averaging of Rotation Matrices References