Online-Learning in Humanoid Robots

Diploma Thesis from the year 2001 in the subject Engineering - Artificial Intelligence, grade: 1,0, Technical University of Berlin (Informatik), language: English, abstract: Inhaltsangabe:Abstract:§Humanoid Robotic Systems have gained an increasing significance in the research world within the last few years. Just five years ago, there were hardly any human-like robots in the world, and those available did not represent human properties at all. They neither looked nor behaved like human beings. Today, a variety of research groups around the world is starting to work on topics related to humanoid robots, and it is very likely that these robots will become important within the upcoming decades even beyond the realm of science.§Trying to determine what humanoid robots are, a first draft of a definition might read as follows: such robots are to be called humanoid robots which - to some extent - are able to live and interact with the everyday human world, and represent certain human features, like cognitive or acting abilities. The main strength of such humanoid robots lies in their ability to operate in surroundings that have been designed for humans in the first place. Humanoid robots can be imagined to become useful assistants for every-day life in areas as diverse as:§- Rescue and clearing of dangerous situations.§- Janitorial services, Housekeeping.§- Security services.§- Care-taking in hospitals, recreational facilities.§- Entertainment.§In all these fields, close human interaction is a core issue and can be regarded as the minimum common basis. The interaction happens on many different levels, from physical touch to gesture recognition and the processing of spoken language. On cognitive issues like the two last named, much research has been done in the past few years. One has, however, to keep in mind that also the physical appearance, e.g. smoothness of motions, is an important issue when designing humanoid robots.§Inhaltsverzeichnis:Table of Contents:§FOREWORD1§1.INTRODUCTION2§1.1INTRODUCING THE AREA OF HUMANOID ROBOTICS2§1.2MECHANICAL DESIGN FOR HUMANOID ROBOTS3§1.3CONTROLLING HUMANOID ROBOTS4§1.4EXAMPLES OF TODAY'S HUMANOID ROBOTS5§1.5THE PURPOSE OF THE THESIS9§2.A BRIEF RECAPITULATION OF BASIC ROBOT CONTROL10§2.1INTRODUCTION TO ROBOT CONTROL10§2.2CONTROLLING THE EXECUTION OF DESIRED TRAJECTORIES12§2.3THE FEEDBACK CONTROL FUNCTION14§2.4THE FEED-FORWARD CONTROL FUNCTION17§2.5ESTIMATING DYNAMICS USING RIGID BODY ASSUMPTIONS18§2.6RECAPITULATION22§2.7CONTROL OF HUMANOID ROBOTS23§3.INTRODUCTION TO ROBOT LEARNING25§3.1GENERAL REMARKS ON ROBOT LEARNING25§3.2THE BIAS / VARIANCE TRADEOFF27§3.3GLOBAL VERSUS LOCAL LEARNING STRATEGIES29§3.4THE CURSE OF DIMENSIONALITY31§3.5ONLINE-LEARNING32§3.6LEARNING INVERSE DYNAMICS34§3.7RESULTS34§4.THE LEARNING ALGORITHM LWPR36§4.1ADVANTAGES OF LEARNING APPROACHES36§4.2DESIRED ALGORITHMIC PROPERTIES38§4.3INPUT DATA PREPROCESSING45§4.4PREDICTING OUTPUT DATA USING LWPR47§4.5LEARNING PARAMETERS50§4.6THE FINAL LWPR ALGORITHM57§4.7PROPERTIES OF LWPR59§5.EVALUATION OF LWPR'S PERFORMANCE ON ARTIFICIAL DATA61§5.1INTRODUCTION61§5.2ONE DIMENSIONAL FUNCTION FITTING61§5.3TWO-DIMENSIONAL FUNCTION FITTING WITH SHIFTING INPUT DISTRIBUTIONS65§5.4LIMITING THE COMPUTATIONAL COMPLEXITY68§5.5DISCUSSION69§6.THE HUMANOID ROBOT71§6.1HARDWARE FOR ROBOT CONTROL71§6.2SOFTWARE FOR ROBOT CONTROL74§6.3THE SARCOS ROBOTIC ARM77§7.EVALUATION OF THE ARM'S MOTION81§7.1GENERATING TRAJECTORIES FOR LEARNING81§7.2LEARNING INVERSE DYNAMICS85§7.3LEARNING RESULTS ON SAMPLED DATA87§7.4VERIFYING THE RESULTS ON REAL MOTION91§8.CONCLUSION96§8.1DISCUSSION96§8.2BRIEF INTRODUCTION TO ...