Artificial Intelligence 6E – George Luger (1)

TR trees offer an intuitive framework for encoding tuning plans acquired from accelerator physicists. In fact with very little help, the physicists themselves were able to develop their own TR trees. Further details of these applications can be found by consulting the references. We next revisit the NASA model-based reasoning example of Section 8.3, to describe control/plan- ning algorithms for the propulsion system of space vehicles.

8.4.4 Planning: a NASA Example (Williams and Nayak) In this section we describe how a planner can be implemented in the context of a model- based reasoner. We continue the NASA example of Williams and Nayak (1996b) intro- duced in Section 8.3.2. Livingstone is a reactive configuration manager that uses a compo- sitional, component-based model of the space craft propulsion system to determine configuration actions, as is seen in Figure 8.23. Each propulsion component is modeled as a transition system that specifies the behaviors of the operating and failure modes of the component, the nominal and failure transitions between modes, and the costs and likelihoods of transitions, as in Figure 8.24.

In Figure 8.24, open and closed are normal operation modes, but stuck open and stuck closed are failure modes. The open command has unit cost and causes a mode transition from closed to open, similarly for the close command. Failure transitions move the valve from normal operating modes to one of the failure modes with probability 0.01.

Mode behaviors are specified using formulae in propositional logic, but transitions between modes are specified using formulae in a restricted temporal, propositional logic. The restricted temporal, propositional logic is adequate for modeling digital hardware, Figure 8.23 Model-based reactive configuration management, from Williams and Nayak (1996b). Planner High-level goals Configuration goals Commands Confirmation Configuration Manager Executive Spacecraft analog hardware using qualitative abstractions (de Kleer and Williams 1989, Weld and de Kleer 1990), and real time software using the models of concurrent reactive systems (Manna and Pnueli 1992).

The spacecraft transition system model is a composition of its component transition systems in which the set of configurations of the spacecraft is the cross product of the sets of component modes. We assume that the component transition systems operate synchronously; that is, for each spacecraft transition every component performs a transition. A model-based configuration manager uses its transition-system model to both iden- tify the current configuration of the spacecraft, called mode estimation ME, and move the spacecraft into a new configuration that achieves the desired configuration goals, called mode reconfiguration, MR. ME incrementally generates all spacecraft transitions from the previous configuration such that the models of the resulting configurations are consistent with the current observations.

This page intentionally left blank Boston San Francisco New York London Toronto Sydney Tokyo Singapore Madrid Mexico City Munich Paris Cape Town Hong Kong Montreal Structures and Strategies for Complex Problem Solving George F Luger University of New Mexico SIXTH EDITION Executive Editor Michael Hirsch Acquisitions Editor Matt Goldstein Editorial Assistant Sarah Milmore Associate Managing Editor Jeffrey Holcomb Digital Assets Manager Marianne Groth Senior Media Producer Bethany Tidd Marketing Manager Erin Davis Senior Author Support/ Technology Specialist Joe Vetere Senior Manufacturing Buyer Carol Melville Text Design, Composition, and Illustrations George F Luger Cover Design Barbara Atkinson Cover Image © Tom Barrow For permission to use copyrighted material, grateful acknowledgment is made to the copyright holders listed on page xv, which is hereby made part of this copyright page.

Many of the designations used by manufacturers and sellers to distinguish their products are claimed as trademarks. Where those designations appear in this book, and Addison-Wesley was aware of a trademark claim, the designations have been printed in initial caps or all caps. Library of Congress Cataloging-in-Publication Data Luger, George F. Artificial intelligence : structures and strategies for complex problem solving / George F. Luger.– 6th ed. p. cm. Includes bibliographical references and index. ISBN-13: 978-0-321-54589-3 (alk. paper) 1. Artificial intelligence. 2. Knowledge representation (Information theory) 3.

Problem solving. 4. PROLOG (Computer program language) 5. LISP (Computer program language) I. Title. Q335.L84 2008 006.3–dc22 2007050376 Copyright © 2009 Pearson Education, Inc. All rights reserved. No part of this publication may be reproduced, stored in a retrieval system, or transmitted, in any form or by any means, electronic, mechanical, photocopying, recording, or otherwise, without the prior written permission of the publisher. Printed in the United States of America. For information on obtaining permission for use of material in this work, please submit a written request to Pearson Education, Inc., Rights and Contracts Department, 501 Boylston Street, Suite 900, Boston, MA 02116, fax (617) 671-3447, or online at http://www.pearsoned.com/legal/permissions.htm.

ISBN-13: 978-0-321-54589-3 ISBN-10: 0-321-54589-3 1 2 3 4 5 6 7 8 9 10—CW—12 11 10 09 08 For my wife, Kathleen, and our children Sarah, David, and Peter. Si quid est in me ingenii, judices . . . Cicero, Pro Archia Poeta GFL This page intentionally left blank PREFACE vii PREFACE What we have to learn to do we learn by doing.

. . —ARISTOTLE, Ethics Welcome to the Sixth Edition! I was very pleased to be asked to produce the sixth edition of my artificial intelligence book. It is a compliment to the earlier editions, started over twenty years ago, that our approach to AI has been so highly valued.

This is a short excerpt from the opening of “” by Unknown, quoted for review and introduction purposes. All rights belong to the copyright holders.

Book Information

• Unique ID: 8ad41e5e9ccb8d55
• File Extension: .pdf
• File Size: 5,354,316 bytes (5.106 MB)
• Title: –
• Author: Unknown
• ISBN: 9780321545893, 0321545893
• Pages: 780
• Language: English (en)

Reading & Word Statistics

• Estimated Reading Time: 1604.64 minutes
• Total Words: 320,928
• Total Characters: 1,973,424
• Average Words per Page: 411.45
• Average Characters per Page: 2530.03

Most Frequent Words

search (1184), problem (1183), state (1144), learning (1103), example (904), figure (883), knowledge (787), set (775), intelligence (712), section (707), chapter (665), system (648), space (603), one (564), also (563), part (556), two (553), rules (551), rule (551), language (537), number (534), systems (528), reasoning (520), algorithm (514), use (510), new (507), used (468), machine (467), representation (465), data (464), goal (454), model (441), network (434), states (415), information (402), using (397), graph (385), possible (384), between (373), first (369), important (354), time (352), function (351), solving (350), predicate (350), probability (337), many (331), see (328), calculus (328), human (327), world (321), path (320), node (315), domain (312), form (303), models (303), value (301), algorithms (299), next (298), natural (297), heuristic (296), problems (292), expert (289), nodes (286), artificial (284), given (284), general (279), solution (273), noun (269), networks (265), logic (263), program (262), process (262), often (260), approach (258), true (255), research (254), sentence (252), based (250), concept (249), well (248), structure (247), computer (245), production (245), understanding (243), values (243), current (242), input (242), different (241), theory (239), examples (239), called (234), thus (231), markov (225), make (225), work (221), verb (221), inference (218), simple (215), genetic (214).

PDF Download