Kawamura / Svinin Advances in Robot Control

From Everyday Physics to Human-Like Movements
1. Auflage 2007
ISBN: 978-3-540-37347-6
Verlag: Springer
Format: PDF
Kopierschutz: 1 - PDF Watermark

E-Book, Englisch, 328 Seiten, eBook

Robotics is still a young science, but we can already identify the people who de?ned its primary course of development. Suguru Arimoto is one of them. His early works laid the foundations of what nowadays is called modern robot control, and we believe it is both appropriate and necessary to write a book on recent advances in this ?eld in the context of his scienti?c interests. While presenting recent advances in robot control is the main intention of this book, we also think it is appropriate to highlight Suguru Arimoto’s research career, main scienti?c achievements, and his personality, too. This can be very inspiring and instructive, especially for young researchers. What are the most remarkable features of Suguru Arimoto? On the p- sonal side, his vitality is striking. He is always focused on a research target, and it is always a fun and a pleasure to discuss with him scienti?c pr- lems and to learn from him. His passion to explain things that might not appear obvious is endless. It is very encouraging to younger researchers that, at this stage of his career, he is still a very active, approachable, and in?u- tial researcher, and a person who leads by example. On the scienti?c side, we should stress his research philosophy. He believes that the ?nal result should be simple and have a clear physical (or physiological, in his recent research) interpretation.

Zielgruppe

Research

Autoren/Hrsg.

Kawamura, Sadao

Svinin, Mikhail

Inhaltsverzeichnis

Human Robotics: A Vision and A Dream.- Human Robotics: A Vision and A Dream.- From Everyday Physics to Robot Control.- Natural Motion and Singularity-Consistent Inversion of Robot Manipulators.- Approximate Jacobian Control for Robot Manipulators.- Adaptive Visual Servoing of Robot Manipulators.- Orthogonalization Principle for Dynamic Visual Servoing of Constrained Robot Manipulators.- Passivity-Based Control of Multi-Agent Systems.- Navigation Functions for Dynamical, Nonholonomically Constrained Mechanical Systems.- Planning and Control of Robot Motion Based on Time-Scale Transformation.- From Robot Control to Human-Like Movements.- Modularity, Synchronization, and What Robotics May Yet Learn from the Brain.- Force Control with A Muscle-Activated Endoskeleton.- On Dynamic Control Mechanisms of Redundant Human Musculo-Skeletal System.- Principle of Superposition in Human Prehension.- Motion Planning of Human-Like Movements in the Manipulation of Flexible Objects.- Haptic Feedback Enhancement Through Adaptive Force Scaling: Theory and Experiment.- Learning to Dynamically Manipulate: A Table Tennis Robot Controls a Ball and Rallies with a Human Being.

Leseproben

Approximate Jacobian Control for Robot Manipulators (P. 35)

Chien Chern Cheah
School of Electrical and Electronic Engineering
Nanyang Technological University
Block S1, Nanyang Avenue, S(639798)
Republic of Singapore

Summary.

Most research so far in robot control has assumed either kinematics or Jacobian matrix of the robots from joint space to task space is known exactly. Unfortunately, no physical parameters can be derived exactly. In addition, when the robot picks up objects of uncertain lengths, orientations or gripping points, the kinematics and dynamics become uncertain and change according to different tasks.

This paper presents several approximate Jacobian control laws for robots with uncertainties in kinematics and dynamics. Lyapunov functions are presented for stability analysis of feedback control problems with uncertain kinematics. We shall show that the end-effector’s position converges to a desired position even when the kinematics and Jacobian matrix are uncertain.

1 Introduction

It is interesting to observe from human reaching movements that we do not need an accurate knowledge of kinematics and dynamics of our arms. We are also able to pick up a new tool or object and manipulate it skillfully to accomplish a task, using only the approximate knowledge of the length, mass, orientation and gripping point of the tool. Such basic ability of sensing and responding to changes without an accurate knowledge of sensory-to-motor transformation gives us a high degree of flexibility in dealing with unforseen changes in the real world.

The kinematics and dynamics of robot manipulators are highly nonlinear. By exploring physical properties of the robot system and using Lyapunov method, Takegaki and Arimoto [1], Arimoto and Miyazaki [2] showed that simple controllers such as the PD and PID feedback are effective for setpoint control despite the nonlinearity and uncertainty of the robot dynamics.

To deal with trajectory-tracking control, Slotine and Li [3, 4] proposed an adaptive control law for robotic manipulator using Lyapunov method. After more than two decades of research, much progress has been made in control of robots with dynamic uncertainty [1]-[19].

However, most research on robot control has assumed that the exact kinematics and Jacobian matrix of the manipulator from joint space to Cartesian space are known. In the presence of uncertainty in kinematics, it is impossible to derive the desired joint angle from the desired end effector path by solving the inverse kinematics problem.

In addition, the Jacobian matrix of the mapping from joint space to task space could not be exactly derived. This assumption leads us to several problems in the development of robot control laws today. In free motion [20], this implies that the exact lengths of the links, joint o.sets and the object which the robot is holding, must be known. Unfortunately, no physical parameters could be derived exactly.

In addition, when the robot picks up objects or tools of di.erent lengths, unknown orientations and gripping points, the overall kinematics are changing and therefore difficult to derive exactly.