Creating Admissible Heuristic Functions: The General Relaxation Principle and Delete Relaxation

Size: px
Start display at page:

Download "Creating Admissible Heuristic Functions: The General Relaxation Principle and Delete Relaxation"

Transcription

1 Creating Admissible Heuristic Functions: The General Relaxation Principle and Delete Relaxation

2 Relaxed Planning Problems 74 A classical planning problem P has a set of solutions Solutions(P) = { π : π is an executable sequence of actions leading from to a state in } Suppose that: P is a classical planning problem P is another classical planning problem Solutions(P) Solutions(P ) If a plan achieves the goal in P, the same plan also achieves the goal in P But P can have additional solutions! Then P is a relaxation of P The constraints on what constitutes a solution have been relaxed

3 Relaxation Example 1 75 Example 1: Adding new actions All old solutions still valid, but new solutions may exist Modifies the state transition system (states/actions) This particular example: shorter solution exists on(b,a) on(a,c) Goal

4 Relaxation Example 2 76 Example 2: Adding goal states All old solutions still valid, but new solutions may exist Retains the same states and transitions This particular example: Optimal solution retains the same length Goal on(b,a) on(a,c) or on(c,b) Goal on(b,a) on(a,c) or on(c,b)

5 Relaxation 77 Definition of relaxation: Change the problem so that all old solutions remain valid, new solutions become possible! An optimal solution for the relaxed problem can never be more expensive than an original optimal solution You ve seen relaxations adapted to the 8-puzzle What can you do independently of the planning domain used?

6 Delete Relaxation (1) 78 One domain-independent technique: delete relaxation Assume a classical problem with: Positive preconditions Positive goals Both negative and positive effects Relaxation: remove all negative effects (all "delete effects")! before transformation: after transformation:

7 No physical meaning! Delete Relaxation (2): Example 79 State space: Original problem State space: delete-relaxed problem Notice: Transitions are not added but moved! Any action applicable (or goal satisfied) after unstack(a,c) in the original problem is applicable (satisfied) after unstack(a,c) here :strips preconds/goals are positive!

8 Delete Relaxation (3): Heuristic 80 So given any action sequence: If it is executable in P, it is executable in delete-relaxed P If it results in a goal state in P, it results in a goal state in del-relaxed P This is a relaxation! Easy to apply mechanically Remove all negative effects ( the delete effects are relaxed/removed ) If only this relaxation is applied: Gives us the optimal delete relaxation heuristic, h+(n) h+(n) = the cost of an optimal solution to a delete-relaxed problem starting in node n Need a heuristic value for every expanded state calculating h+(n) still takes too much time (requires planning)!

9 Important Issues 81 Important: You cannot use a relaxed problem as a heuristic. What would that mean? You use the cost of an optimal solution to the relaxed problem as a heuristic. Solving the relaxed problem can result in a more expensive solution inadmissible! You have to solve it optimally to get the admissibility guarantee. You don t just solve the relaxed problem once. Every time you reach a new state and want to calculate a heuristic, you have to solve the relaxed problem of getting from that state to the goal.

10 The h 1 Heuristic

11 The h 1 Heuristic 83 We need a faster heuristic! How about relaxing h+ further? ) (optimal delete relaxation): Remove delete effects, find a long plan achieving all goals Find a separate plan for each goal fact, in a way that ignores delete effects Take the maximum of their costs Size: Grows quickly with plan length Much easier/faster, since search trees tend to be wide!

12 The h 1 Heuristic: Example 84 Cheaper! We discover this is expensive

13 The h 1 Heuristic: Important Property 85 Each goal considered separately! This is why it is fast! No need to consider interactions no combinatorial explosion Each precondition considered separately! Can generalize, consider all pairs of goals/preconds Yields the h 2 heuristic, more powerful, still reasonably fast Each precondition considered separately!

14 Search Strategies for Admissible Heuristics 86 Given an admissible heuristic, we often use A* (not always) Guarantees optimal plans, just like Dijkstra Comparison: Let len(p) be the number of actions in plan p Let h(p) be an admissible heuristic (applied to the final state of plan p) Blocks world, blocks initially on the table, goal is a -block tower 399 blocks to pick up and stack Dijkstra considers plans where len(p) <= 798 A* considers plans where len(p) + h(p) <= 798 len(p)=400 h(p)=398 Explored by Dijkstra, not by A* len(p)=401 h(p)=398 len(p)=401 h(p)=397

15 Admissible Heuristics 87 Admissible heuristics: Must never overestimate To guarantee this, they often underestimate (by a large margin)! Example: h1 -- could have hundreds of goals but we can only count the most expensive one! Achieving clear(d) seems exactly as expensive as achieving all of the goals above Admissible heuristics help, but we d prefer to cut off search even earlier!

16 Non-admissible Heuristics for Non-Optimal Forward State Space Planning

17 Optimal and Satisficing Planning 89 Do we need admissibility? If we must guarantee optimal plans: Need optimal search strategies such as A* Must use admissible heuristics If we are satisfied with a high-quality plan: Satisficing planning Find a plan that is sufficiently good, sufficiently quickly Can use non-admissible heuristics which can often be more informative (give us more help in finding reasonably good alternatives) Generally much faster Handles more complex domains, larger problem instances

18 h add : Use sum instead of maximum 90 Cheaper! Turns out to be expensive

19 The h add Heuristic: Properties 91 The h add heuristic: Can under- and over-estimate: Not admissible! The same action can be counted twice Can miss the opportunity to achieve two goals with one action but we could have used a single teardownalltowers action! But: Retains much more information Example: Every unsatisfied goal contributes to the heuristic value + = 5

20 Search Algorithms for h add 92 Which search algorithm to use? A* expands many nodes (expensive), and for h add, there is still no optimality guarantee! HSP planner, and some others, use Hill Climbing Successors: States created by applying an action to the current state Value: Calculated as h add (maximize h add minimize +h add ) Only considers children of the current node tries to move quickly towards low heuristic values doesn t examine all nodes with len(p)+f(p) <= some value

21 Local Optima in Planning Problem: 93 Can end up in a local maximum (where h() is locally minimal) This is not good enough : The goal is not satisfied! Possible solution: Restart search from another expanded state

22 Local Optima: Restarting Init 94 Local Optimum on(b,a) on(a,c) Goal Restart in any of the gray nodes

23 Search Algorithms for h add 95 Late s: State-space planning too simple to be efficient! Most planners used very elaborate and complex search methods Planning competition : HSP (hill climbing + h add + tweaks) solved more problems than most other planners Now, heuristic forward search is the predominant approach!

24

25 Backward Search 97 Classical Planning: Find a path in a finite graph initial Where would we have to be before? What about this direction: Backward, goal initial? goal goal goal What could the previous action be? If we want to be here

26 Simple Backward Search (1) 98 Simple case: A single, completely defined goal state! This must have been true before The goal is not already achieved was executed This must have been true before was executed

27 Simple Backward Search (2) 99 Simple case: A single, completely defined goal state! Relevant actions: Now these states are goals

28 Simple Backward Search (3) 100 Analysis But we don t know how to reach them, or even if it is possible! If we reach one of these states, we know how to continue to the goal! We still need guidance: Heuristics, or something else

29

30 Reachable States 102 At every step, there is a set of states we can reach Example: Blocks World with blocks A/B Step 0: Step 1: Step 2: Step 3: s0 s1, s2 s0, s3, s4 (not s1 or s2!)

31 Reachable States (2) 103 Example: Blocks World, 3 blocks Step 0: s0 Step 1: s1, s2, s3 Step 2: s0, s4, s5, s10, s11, s16, s17 Would take a lot of time to calculate!

32 Possibly Reachable States (1) 104 We can quickly estimate the sets of reachable states Contains all truly reachable states, and some others (exact calculations are too costly!) Goal not satisfied Goal not satisfied in any state Goal not satisfied in any state Goal satisfied in The goal might be achievable using actions!

33 Possibly Reachable States (2) 105 We can use this to shrink the backward search space The goal seems to be reachable here, but we can t be sure (overestimating!)

34 Possibly Reachable States (3) 106 Backward search can then verify what is truly reachable In a much smaller search space than plain backward search The goal was not reachable here! Also not reachable! These are all of the relevant actions none is applicable at time 2!

35 Possibly Reachable States (4) Extend one time step and try again A larger number of states may be truly reachable then 107 Time 4 Continue backward search (complete search procedure with backtracking, etc) Use possibly reachable states to prune the search tree smaller search space

36 Planning graph! GraphPlan: Planning Graph 108 This is the idea behind GraphPlan! Also keeps track of possibly executable actions Proposition levels Which facts might possibly be true in each step? Details are in the book! Initial state 10 possibly executable actions 47 possibly executable actions 1200 possibly executable actions Action levels Which actions may possibly be executed in each step?

37 GraphPlan and Heuristics 109 When GraphPlan was introduced: Very successful, surprisingly fast Current use of planning graphs: The basis of very successful (non-admissible) heuristics, more informative than h add! Calculate heuristic value for state s: Apply delete relaxation Create planning graph starting in state s Backward search finds a solution to the relaxed problem No delete effects smaller graph, simpler calculations No delete effects smaller graph, no backtracking Cost of solution = heuristic value!

38

39 Plan-Space Search 111 Just because we are looking for a path in this graph initial goal goal goal doesn t mean we have to use it as a search space! What if a search step could insert actions anywhere in a plan?

40 initaction goalaction Partial-Order Planning 112 Partial-order planning starts with: A fake initial action, whose effects are the facts of the initial state A fake goal action, whose preconditions are the conjunctive goal facts clear(a) ontable(a) clear(b) ontable(b) clear(a) on(a,b) on(b,c) clear(c) on(c,d) ontable(d) handempty A set of flaws! First type: Open goals to be achieved on(c,d) ontable(d) handempty An ordering relation between these actions

41 initaction goalaction Resolving Open Goals To resolve an open goal: Alternative 1: Add a causal link from an action that already achieves it clear(a) ontable(a) clear(b) ontable(b) clear(a) on(a,b) on(b,c) clear(c) on(c,d) ontable(d) handempty on(c,d) ontable(d) handempty

42 initaction goalaction stack(a,b) Resolving Open Goals To resolve an open goal: holding(a) clear(b) holding(a) clear(b) Alternative 2: Add a new action with the open goal as an effect clear(a) on(a,b) clear(a) ontable(a) on(a,b) clear(b) ontable(b) handempty clear(a) on(b,c) clear(c) on(c,d) on(c,d) ontable(d) ontable(d) handempty handempty

43 pickup(b) stack(b,c) initaction goalaction stack(a,b) Partial Orders 115 Some pairs of actions can remain unordered holding(a) holding(a) clear(b) clear(b) clear(a) on(a,b) clear(a) ontable(a) on(a,b) clear(b) handempty on(b,c) ontable(b) clear(a) clear(c) on(c,d) ontable(d) handempty handempty clear(b) ontable(b) holding(b) clear(c) on(b,c) on(c,d) ontable(d) handempty handempty holding(b) holding(b) handempty clear(b) clear(c) clear(b)

44 pickup(b) stack(b,c) initaction goalaction stack(a,b) Threats 116 Second flaw type: A threat to be resolved initaction supports clear(b) for stack(a,b) there s a causal link pickup(b) deletes clear(b), and may occur between initaction and stack(a,b) We can t be certain that clear(b) still holds when stack(a,b) starts! holding(a) holding(a) clear(b) clear(b) clear(a) on(a,b) clear(a) ontable(a) on(a,b) clear(b) handempty on(b,c) ontable(b) clear(a) clear(c) on(c,d) ontable(d) handempty handempty clear(b) ontable(b) holding(b) clear(c) on(b,c) on(c,d) ontable(d) handempty handempty holding(b) holding(b) handempty clear(b) clear(c) clear(b)

45 Resolving Threats How to make sure that holds when starts? Alternative : The action that disturbs the precondition is placed after the action that has the precondition Only possible if the resulting partial order is consistent (acyclic)!

46 Resolving Threats Alternative 2: The action that disturbs the precondition is placed before the action that supports the precondition Only possible if the resulting partial order is consistent not in this case!

47 Heuristics 119 Again, this is not sufficient in itself Heuristics are required But it has advantages No need to commit to order

48

49 Example Questions 121 Example questions: Describe: Relaxation Delete relaxation The optimal relaxation heuristic The different kinds of flaws that can occur in a partially ordered plan Why: Don t planners use Dijkstra s algorithm to search the state space? Can the h_1 heuristic be calculated more quickly than the optimal delete relaxation heuristic h+? Is the number of open goals an admissible heuristic? Why / why not? Describe one important difference between backward and forward search Given a domain, problem and initial state: Expand the forward state space by two levels Expand the backward state space by two levels

50 TDDD48 Automated Planning (1) 122 Formal basis for planning Alternative representations of planning problems Simple and complex state transition systems Deeper discussions about all of these topics Completely different principles for heuristics Landmarks Pattern databases Alternative search spaces Planning by conversion to boolean satisfiability problems Extended expressivity Planning with time and concurrency Planning with probabilistic actions (Markov Decision Processes) Planning with non-classical goals

51 TDDD48 Automated Planning (2) Example: concurrent actions with duration 123 Duration dependent on parameters, current state Conditions / effects at the start or end of the action Implemented by quite a few planners affects algorithms and heuristics!

52 TDDD48 Automated Planning (3) 124 Planning with domain knowledge Using what you know: Temporal control rules Breaking down a task into smaller parts: Hierarchical Task Networks The travel task has a method called air-travel travel(x,y) air-travel(x,y) Task Method buy-ticket (airport(x), airport(y)) travel (x, airport(x)) fly(airport(x), airport(y)) travel (airport(y), y) Alternative types of planning Path planning And so on

Planning Algorithms Properties Soundness

Planning Algorithms Properties Soundness Chapter MK:VI III. Planning Agent Systems Deductive Reasoning Agents Planning Language Planning Algorithms State-Space Planning Plan-Space Planning HTN Planning Complexity of Planning Problems Extensions

More information

Primitive goal based ideas

Primitive goal based ideas Primitive goal based ideas Once you have the gold, your goal is to get back home s Holding( Gold, s) GoalLocation([1,1], s) How to work out actions to achieve the goal? Inference: Lots more axioms. Explodes.

More information

Fast Forward Planning. By J. Hoffmann and B. Nebel

Fast Forward Planning. By J. Hoffmann and B. Nebel Fast Forward Planning By J. Hoffmann and B. Nebel Chronology faster Traditional Optimal Sub-optimal ««1995 Graphplan-based Satplan-based 2000 Heuristic-based Fast Forward (FF) Winner of AIPS2000 Forward-chaining

More information

Automated Planning. Plan-Space Planning / Partial Order Causal Link Planning

Automated Planning. Plan-Space Planning / Partial Order Causal Link Planning Automated Planning Plan-Space Planning / Partial Order Causal Link Planning Jonas Kvarnström Automated Planning Group Department of Computer and Information Science Linköping University Partly adapted

More information

Planning, Execution & Learning 1. Heuristic Search Planning

Planning, Execution & Learning 1. Heuristic Search Planning Planning, Execution & Learning 1. Heuristic Search Planning Reid Simmons Planning, Execution & Learning: Heuristic 1 Simmons, Veloso : Fall 2001 Basic Idea Heuristic Search Planning Automatically Analyze

More information

CSE 3402: Intro to Artificial Intelligence Planning

CSE 3402: Intro to Artificial Intelligence Planning CSE 3402: Intro to Artificial Intelligence Planning Readings: Russell & Norvig 3 rd edition Chapter 10 (in 2 nd edition, Sections 11.1, 11.2, and 11.4) 1 CWA Classical Planning. No incomplete or uncertain

More information

Classical Planning. CS 486/686: Introduction to Artificial Intelligence Winter 2016

Classical Planning. CS 486/686: Introduction to Artificial Intelligence Winter 2016 Classical Planning CS 486/686: Introduction to Artificial Intelligence Winter 2016 1 Classical Planning A plan is a collection of actions for performing some task (reaching some goal) If we have a robot

More information

Set 9: Planning Classical Planning Systems. ICS 271 Fall 2013

Set 9: Planning Classical Planning Systems. ICS 271 Fall 2013 Set 9: Planning Classical Planning Systems ICS 271 Fall 2013 Outline: Planning Classical Planning: Situation calculus PDDL: Planning domain definition language STRIPS Planning Planning graphs Readings:

More information

Artificial Intelligence

Artificial Intelligence Artificial Intelligence Lecturer 7 - Planning Lecturer: Truong Tuan Anh HCMUT - CSE 1 Outline Planning problem State-space search Partial-order planning Planning graphs Planning with propositional logic

More information

Plan Generation Classical Planning

Plan Generation Classical Planning Plan Generation Classical Planning Manuela Veloso Carnegie Mellon University School of Computer Science 15-887 Planning, Execution, and Learning Fall 2016 Outline What is a State and Goal What is an Action

More information

CS 621 Artificial Intelligence. Lecture 31-25/10/05. Prof. Pushpak Bhattacharyya. Planning

CS 621 Artificial Intelligence. Lecture 31-25/10/05. Prof. Pushpak Bhattacharyya. Planning CS 621 Artificial Intelligence Lecture 31-25/10/05 Prof. Pushpak Bhattacharyya Planning 1 Planning Definition : Planning is arranging a sequence of actions to achieve a goal. Uses core areas of AI like

More information

Planning (under complete Knowledge)

Planning (under complete Knowledge) CSC384: Intro to Artificial Intelligence Planning (under complete Knowledge) We cover Sections11.1, 11.2, 11.4. Section 11.3 talks about partial-order planning, an interesting idea that hasn t really done

More information

CPS 270: Artificial Intelligence Planning

CPS 270: Artificial Intelligence   Planning CPS 270: Artificial Intelligence http://www.cs.duke.edu/courses/fall08/cps270/ Planning Instructor: Vincent Conitzer Planning We studied how to take actions in the world (search) We studied how to represent

More information

Artificial Intelligence Planning

Artificial Intelligence Planning Artificial Intelligence Planning Instructor: Vincent Conitzer Planning We studied how to take actions in the world (search) We studied how to represent objects, relations, etc. (logic) Now we will combine

More information

Domain-Configurable Planning: Hierarchical Task Networks

Domain-Configurable Planning: Hierarchical Task Networks Automated Planning Domain-Configurable Planning: Hierarchical Task Networks Jonas Kvarnström Automated Planning Group Department of Computer and Information Science Linköping University jonas.kvarnstrom@liu.se

More information

Planning. Some material taken from D. Lin, J-C Latombe

Planning. Some material taken from D. Lin, J-C Latombe RN, Chapter 11 Planning Some material taken from D. Lin, J-C Latombe 1 Logical Agents Reasoning [Ch 6] Propositional Logic [Ch 7] Predicate Calculus Representation [Ch 8] Inference [Ch 9] Implemented Systems

More information

Set 9: Planning Classical Planning Systems. ICS 271 Fall 2014

Set 9: Planning Classical Planning Systems. ICS 271 Fall 2014 Set 9: Planning Classical Planning Systems ICS 271 Fall 2014 Planning environments Classical Planning: Outline: Planning Situation calculus PDDL: Planning domain definition language STRIPS Planning Planning

More information

A deterministic action is a partial function from states to states. It is partial because not every action can be carried out in every state

A deterministic action is a partial function from states to states. It is partial because not every action can be carried out in every state CmSc310 Artificial Intelligence Classical Planning 1. Introduction Planning is about how an agent achieves its goals. To achieve anything but the simplest goals, an agent must reason about its future.

More information

Intelligent Systems. Planning. Copyright 2010 Dieter Fensel and Ioan Toma

Intelligent Systems. Planning. Copyright 2010 Dieter Fensel and Ioan Toma Intelligent Systems Planning Copyright 2010 Dieter Fensel and Ioan Toma 1 Where are we? # Title 1 Introduction 2 Propositional Logic 3 Predicate Logic 4 Reasoning 5 Search Methods 6 CommonKADS 7 Problem-Solving

More information

Scale-up Revolution in Automated Planning

Scale-up Revolution in Automated Planning Scale-up Revolution in Automated Planning (Or 1001 Ways to Skin a Planning Graph for Heuristic Fun & Profit) Subbarao Kambhampati http://rakaposhi.eas.asu.edu RIT; 2/17/06 With thanks to all the Yochan

More information

Planning. Chapter 10 of the textbook. Acknowledgement: Thanks to Sheila McIlraith for the slides.

Planning. Chapter 10 of the textbook. Acknowledgement: Thanks to Sheila McIlraith for the slides. Planning Chapter 10 of the textbook. Acknowledgement: Thanks to Sheila McIlraith for the slides. 1 plan n. 1. A scheme, program, or method worked out beforehand for the accomplishment of an objective:

More information

Seach algorithms The travelling salesman problem The Towers of Hanoi Playing games. Comp24412: Symbolic AI. Lecture 4: Search. Ian Pratt-Hartmann

Seach algorithms The travelling salesman problem The Towers of Hanoi Playing games. Comp24412: Symbolic AI. Lecture 4: Search. Ian Pratt-Hartmann Comp24412: Symbolic AI Lecture 4: Search Ian Pratt-Hartmann Room KB2.38: email: ipratt@cs.man.ac.uk 2016 17 Outline Seach algorithms The travelling salesman problem The Towers of Hanoi Playing games Typical

More information

Planning Chapter

Planning Chapter Planning Chapter 11.1-11.3 Some material adopted from notes by Andreas Geyer-Schulz and Chuck Dyer Typical BW planning problem A C B A B C Blocks world The blocks world is a micro-world consisting of a

More information

CMU-Q Lecture 6: Planning Graph GRAPHPLAN. Teacher: Gianni A. Di Caro

CMU-Q Lecture 6: Planning Graph GRAPHPLAN. Teacher: Gianni A. Di Caro CMU-Q 15-381 Lecture 6: Planning Graph GRAPHPLAN Teacher: Gianni A. Di Caro PLANNING GRAPHS Graph-based data structure representing a polynomial-size/time approximation of the exponential search tree Can

More information

Principles of AI Planning. Principles of AI Planning. 8.1 Parallel plans. 8.2 Relaxed planning graphs. 8.3 Relaxation heuristics. 8.

Principles of AI Planning. Principles of AI Planning. 8.1 Parallel plans. 8.2 Relaxed planning graphs. 8.3 Relaxation heuristics. 8. Principles of AI Planning June th, 8. Planning as search: relaxation heuristics Principles of AI Planning 8. Planning as search: relaxation heuristics alte Helmert and Bernhard Nebel Albert-Ludwigs-Universität

More information

1 What is Planning? automatic programming. cis716-spring2004-parsons-lect17 2

1 What is Planning? automatic programming. cis716-spring2004-parsons-lect17 2 PLANNING 1 What is Planning? Key problem facing agent is deciding what to do. We want agents to be taskable: give them goals to achieve, have them decide for themselves how to achieve them. Basic idea

More information

Commitment Least you haven't decided where to go shopping. Or...suppose You can get milk at the convenience store, at the dairy, or at the supermarket

Commitment Least you haven't decided where to go shopping. Or...suppose You can get milk at the convenience store, at the dairy, or at the supermarket Planning as Search-based Problem Solving? Imagine a supermarket shopping scenario using search-based problem solving: Goal: buy milk and bananas Operator: buy Heuristic function: does = milk

More information

Intelligent Agents. State-Space Planning. Ute Schmid. Cognitive Systems, Applied Computer Science, Bamberg University. last change: 14.

Intelligent Agents. State-Space Planning. Ute Schmid. Cognitive Systems, Applied Computer Science, Bamberg University. last change: 14. Intelligent Agents State-Space Planning Ute Schmid Cognitive Systems, Applied Computer Science, Bamberg University last change: 14. April 2016 U. Schmid (CogSys) Intelligent Agents last change: 14. April

More information

BASIC PLAN-GENERATING SYSTEMS

BASIC PLAN-GENERATING SYSTEMS CHAPTER 7 BASIC PLAN-GENERATING SYSTEMS In chapters 5 and 6 we saw that a wide class of deduction tasks could be solved by commutative production systems. For many other problems of interest in AI, however,

More information

Where are we? Informatics 2D Reasoning and Agents Semester 2, Planning with state-space search. Planning with state-space search

Where are we? Informatics 2D Reasoning and Agents Semester 2, Planning with state-space search. Planning with state-space search Informatics 2D Reasoning and Agents Semester 2, 2018 2019 Alex Lascarides alex@inf.ed.ac.uk Where are we? Last time... we defined the planning problem discussed problem with using search and logic in planning

More information

Planning. Introduction

Planning. Introduction Planning Introduction Planning vs. Problem-Solving Representation in Planning Systems Situation Calculus The Frame Problem STRIPS representation language Blocks World Planning with State-Space Search Progression

More information

Efficient Planning with State Trajectory Constraints

Efficient Planning with State Trajectory Constraints Efficient Planning with State Trajectory Constraints Stefan Edelkamp Baroper Straße 301 University of Dortmund email: stefan.edelkamp@cs.uni-dortmund.de Abstract. This paper introduces a general planning

More information

Does Representation Matter in the Planning Competition?

Does Representation Matter in the Planning Competition? Does Representation Matter in the Planning Competition? Patricia J. Riddle Department of Computer Science University of Auckland Auckland, New Zealand (pat@cs.auckland.ac.nz) Robert C. Holte Computing

More information

Today s s lecture. Lecture 3: Search - 2. Problem Solving by Search. Agent vs. Conventional AI View. Victor R. Lesser. CMPSCI 683 Fall 2004

Today s s lecture. Lecture 3: Search - 2. Problem Solving by Search. Agent vs. Conventional AI View. Victor R. Lesser. CMPSCI 683 Fall 2004 Today s s lecture Search and Agents Material at the end of last lecture Lecture 3: Search - 2 Victor R. Lesser CMPSCI 683 Fall 2004 Continuation of Simple Search The use of background knowledge to accelerate

More information

A* optimality proof, cycle checking

A* optimality proof, cycle checking A* optimality proof, cycle checking CPSC 322 Search 5 Textbook 3.6 and 3.7.1 January 21, 2011 Taught by Mike Chiang Lecture Overview Recap Admissibility of A* Cycle checking and multiple path pruning Slide

More information

Acknowledgements. Outline

Acknowledgements. Outline Acknowledgements Heuristic Search for Planning Sheila McIlraith University of Toronto Fall 2010 Many of the slides used in today s lecture are modifications of slides developed by Malte Helmert, Bernhard

More information

Artificial Intelligence Planning

Artificial Intelligence Planning Artificial Intelligence Planning Gholamreza Ghassem-Sani Sharif University of Technology Lecture slides are mainly based on Dana Nau s AI Planning University of Maryland 1 For technical details: Related

More information

Artificial Intelligence. Planning

Artificial Intelligence. Planning Artificial Intelligence Planning Planning Planning agent Similar to previous problem solving agents Constructs plans that achieve its goals, then executes them Differs in way it represents and searches

More information

AUTOMATED PLANNING Automated Planning sequence of actions goal

AUTOMATED PLANNING Automated Planning sequence of actions goal AUTOMATED PLANNING Automated Planning is an important problem solving activity which consists in synthesizing a sequence of actions performed by an agent that leads from an initial state of the world to

More information

Artificial Intelligence 2004 Planning: Situation Calculus

Artificial Intelligence 2004 Planning: Situation Calculus 74.419 Artificial Intelligence 2004 Planning: Situation Calculus Review STRIPS POP Hierarchical Planning Situation Calculus (John McCarthy) situations actions axioms Review Planning 1 STRIPS (Nils J. Nilsson)

More information

4 INFORMED SEARCH AND EXPLORATION. 4.1 Heuristic Search Strategies

4 INFORMED SEARCH AND EXPLORATION. 4.1 Heuristic Search Strategies 55 4 INFORMED SEARCH AND EXPLORATION We now consider informed search that uses problem-specific knowledge beyond the definition of the problem itself This information helps to find solutions more efficiently

More information

Hill Climbing. Assume a heuristic value for each assignment of values to all variables. Maintain an assignment of a value to each variable.

Hill Climbing. Assume a heuristic value for each assignment of values to all variables. Maintain an assignment of a value to each variable. Hill Climbing Many search spaces are too big for systematic search. A useful method in practice for some consistency and optimization problems is hill climbing: Assume a heuristic value for each assignment

More information

Automated Planning & Artificial Intelligence

Automated Planning & Artificial Intelligence Automated Planning & Artificial Intelligence Uninformed and Informed search in state space Humbert Fiorino Humbert.Fiorino@imag.fr http://membres-lig.imag.fr/fiorino Laboratory of Informatics of Grenoble

More information

Planning plan Synonyms: plan Synonyms:

Planning plan Synonyms: plan Synonyms: Planning This material is covered in R&N 2 nd edition chapters 10.3, 11.1, 11.2, 11.4 This material is covered in R&N 3 rd edition chapter 10 plan n. 1. A scheme, program, or method worked out beforehand

More information

Artificial Intelligence

Artificial Intelligence Artificial Intelligence Exercises & Solutions Chapters -4: Search methods. Search Tree. Draw the complete search tree (starting from S and ending at G) of the graph below. The numbers beside the nodes

More information

EMPLOYING DOMAIN KNOWLEDGE TO IMPROVE AI PLANNING EFFICIENCY *

EMPLOYING DOMAIN KNOWLEDGE TO IMPROVE AI PLANNING EFFICIENCY * Iranian Journal of Science & Technology, Transaction B, Engineering, Vol. 29, No. B1 Printed in The Islamic Republic of Iran, 2005 Shiraz University EMPLOYING DOMAIN KNOWLEDGE TO IMPROVE AI PLANNING EFFICIENCY

More information

DIT411/TIN175, Artificial Intelligence. Peter Ljunglöf. 23 January, 2018

DIT411/TIN175, Artificial Intelligence. Peter Ljunglöf. 23 January, 2018 DIT411/TIN175, Artificial Intelligence Chapters 3 4: More search algorithms CHAPTERS 3 4: MORE SEARCH ALGORITHMS DIT411/TIN175, Artificial Intelligence Peter Ljunglöf 23 January, 2018 1 TABLE OF CONTENTS

More information

Heuristic Search. Heuristic Search. Heuristic Search. CSE 3401: Intro to AI & LP Informed Search

Heuristic Search. Heuristic Search. Heuristic Search. CSE 3401: Intro to AI & LP Informed Search CSE 3401: Intro to AI & LP Informed Search Heuristic Search. Required Readings: Chapter 3, Sections 5 and 6, and Chapter 4, Section 1. In uninformed search, we don t try to evaluate which of the nodes

More information

STRIPS HW 1: Blocks World

STRIPS HW 1: Blocks World STRIPS HW 1: Blocks World Operator Precondition Delete List Add List Stack(x, y) CLEAR(y) CLEAR(y) HOLDING(x) HOLDING(x) ON(x, y) Unstack(x, y) ON(x, y) ON(x, y) HOLDING(x) CLEAR(x) CLEAR(y) PickUp(x)

More information

Informed (Heuristic) Search. Idea: be smart about what paths to try.

Informed (Heuristic) Search. Idea: be smart about what paths to try. Informed (Heuristic) Search Idea: be smart about what paths to try. 1 Blind Search vs. Informed Search What s the difference? How do we formally specify this? A node is selected for expansion based on

More information

Chapter 1 Introduc0on

Chapter 1 Introduc0on Lecture slides for Automated Planning: Theory and Practice Chapter 1 Introduc0on Dana S. Nau University of Maryland Updated 3:03 PM December 14, 2013 1 Some Dic0onary Defini0ons of Plan plan n. 1. A scheme,

More information

Informed Search CS457 David Kauchak Fall 2011

Informed Search CS457 David Kauchak Fall 2011 Admin Informed Search CS57 David Kauchak Fall 011 Some material used from : Sara Owsley Sood and others Q3 mean: 6. median: 7 Final projects proposals looked pretty good start working plan out exactly

More information

Planning (What to do next?) (What to do next?)

Planning (What to do next?) (What to do next?) Planning (What to do next?) (What to do next?) (What to do next?) (What to do next?) (What to do next?) (What to do next?) CSC3203 - AI in Games 2 Level 12: Planning YOUR MISSION learn about how to create

More information

Planning. CPS 570 Ron Parr. Some Actual Planning Applications. Used to fulfill mission objectives in Nasa s Deep Space One (Remote Agent)

Planning. CPS 570 Ron Parr. Some Actual Planning Applications. Used to fulfill mission objectives in Nasa s Deep Space One (Remote Agent) Planning CPS 570 Ron Parr Some Actual Planning Applications Used to fulfill mission objectives in Nasa s Deep Space One (Remote Agent) Particularly important for space operations due to latency Also used

More information

Planning and Acting. CITS3001 Algorithms, Agents and Artificial Intelligence. 2018, Semester 2

Planning and Acting. CITS3001 Algorithms, Agents and Artificial Intelligence. 2018, Semester 2 Planning and Acting CITS3001 Algorithms, Agents and Artificial Intelligence Tim French School of Computer Science and Software Engineering The University of Western Australia 2018, Semester 2 Summary We

More information

3.6.2 Generating admissible heuristics from relaxed problems

3.6.2 Generating admissible heuristics from relaxed problems 3.6.2 Generating admissible heuristics from relaxed problems To come up with heuristic functions one can study relaxed problems from which some restrictions of the original problem have been removed The

More information

Heuristic Search for Planning

Heuristic Search for Planning Heuristic Search for Planning Sheila McIlraith University of Toronto Fall 2010 S. McIlraith Heuristic Search for Planning 1 / 50 Acknowledgements Many of the slides used in today s lecture are modifications

More information

Chapter 3: Search. c D. Poole, A. Mackworth 2010, W. Menzel 2015 Artificial Intelligence, Chapter 3, Page 1

Chapter 3: Search. c D. Poole, A. Mackworth 2010, W. Menzel 2015 Artificial Intelligence, Chapter 3, Page 1 Chapter 3: Search c D. Poole, A. Mackworth 2010, W. Menzel 2015 Artificial Intelligence, Chapter 3, Page 1 Searching Often we are not given an algorithm to solve a problem, but only a specification of

More information

Principles of AI Planning

Principles of AI Planning Principles of AI Planning 8. Planning as search: relaxation Albert-Ludwigs-Universität Freiburg Bernhard Nebel and Robert Mattmüller November 9th, 4 Plan steps Plan steps, serializations and parallel Forward

More information

Planning. Introduction

Planning. Introduction Introduction vs. Problem-Solving Representation in Systems Situation Calculus The Frame Problem STRIPS representation language Blocks World with State-Space Search Progression Algorithms Regression Algorithms

More information

! A* is best first search with f(n) = g(n) + h(n) ! Three changes make it an anytime algorithm:

! A* is best first search with f(n) = g(n) + h(n) ! Three changes make it an anytime algorithm: Anytime A* Today s s lecture Lecture 5: Search - 4 Hierarchical A* Jiaying Shen CMPSCI 683 Fall 2004 Other Examples of Hierarchical Problem Solving Reviews of A* and its varations 2 Anytime algorithms

More information

Outline. Best-first search

Outline. Best-first search Outline Best-first search Greedy best-first search A* search Heuristics Admissible Heuristics Graph Search Consistent Heuristics Local search algorithms Hill-climbing search Beam search Simulated annealing

More information

Informed search algorithms. (Based on slides by Oren Etzioni, Stuart Russell)

Informed search algorithms. (Based on slides by Oren Etzioni, Stuart Russell) Informed search algorithms (Based on slides by Oren Etzioni, Stuart Russell) The problem # Unique board configurations in search space 8-puzzle 9! = 362880 15-puzzle 16! = 20922789888000 10 13 24-puzzle

More information

DFS. Depth-limited Search

DFS. Depth-limited Search DFS Completeness? No, fails in infinite depth spaces or spaces with loops Yes, assuming state space finite. Time complexity? O(b m ), terrible if m is much bigger than d. can do well if lots of goals Space

More information

Validating Plans with Durative Actions via Integrating Boolean and Numerical Constraints

Validating Plans with Durative Actions via Integrating Boolean and Numerical Constraints Validating Plans with Durative Actions via Integrating Boolean and Numerical Constraints Roman Barták Charles University in Prague, Faculty of Mathematics and Physics Institute for Theoretical Computer

More information

Heuristic Search and Advanced Methods

Heuristic Search and Advanced Methods Heuristic Search and Advanced Methods Computer Science cpsc322, Lecture 3 (Textbook Chpt 3.6 3.7) May, 15, 2012 CPSC 322, Lecture 3 Slide 1 Course Announcements Posted on WebCT Assignment1 (due on Thurs!)

More information

Advanced A* Improvements

Advanced A* Improvements Advanced A* Improvements 1 Iterative Deepening A* (IDA*) Idea: Reduce memory requirement of A* by applying cutoff on values of f Consistent heuristic function h Algorithm IDA*: 1. Initialize cutoff to

More information

Recap A Search Optimality of A. Search: A. CPSC 322 Search 5. Textbook 3.6. Search: A CPSC 322 Search 5, Slide 1

Recap A Search Optimality of A. Search: A. CPSC 322 Search 5. Textbook 3.6. Search: A CPSC 322 Search 5, Slide 1 Search: A CPSC 322 Search 5 Textbook 3.6 Search: A CPSC 322 Search 5, Slide 1 Lecture Overview 1 Recap 2 A Search 3 Optimality of A Search: A CPSC 322 Search 5, Slide 2 Search with Costs Sometimes there

More information

EE562 ARTIFICIAL INTELLIGENCE FOR ENGINEERS

EE562 ARTIFICIAL INTELLIGENCE FOR ENGINEERS EE562 ARTIFICIAL INTELLIGENCE FOR ENGINEERS Lecture 4, 4/11/2005 University of Washington, Department of Electrical Engineering Spring 2005 Instructor: Professor Jeff A. Bilmes Today: Informed search algorithms

More information

Section 9: Planning CPSC Artificial Intelligence Michael M. Richter

Section 9: Planning CPSC Artificial Intelligence Michael M. Richter Section 9: Planning Actions (1) An action changes a situation and their description has to describe this change. A situation description (or a state) sit is a set of ground atomic formulas (sometimes literals

More information

CS 512: Comments on Graph Search 16:198:512 Instructor: Wes Cowan

CS 512: Comments on Graph Search 16:198:512 Instructor: Wes Cowan CS 512: Comments on Graph Search 16:198:512 Instructor: Wes Cowan 1 General Graph Search In general terms, the generic graph search algorithm looks like the following: def GenerateGraphSearchTree(G, root):

More information

MASSACHUSETTS INSTITUTE OF TECHNOLOGY Department of Electrical Engineering and Computer Science Artificial Intelligence Fall, 2010

MASSACHUSETTS INSTITUTE OF TECHNOLOGY Department of Electrical Engineering and Computer Science Artificial Intelligence Fall, 2010 MSSHUSETTS INSTITUTE OF TEHNOLOY epartment of Electrical Engineering and omputer Science 6.0 rtificial Intelligence Fall, 00 Search Me! Recitation, Thursday September Prof. ob erwick. ifference between

More information

Distributed Algorithms 6.046J, Spring, Nancy Lynch

Distributed Algorithms 6.046J, Spring, Nancy Lynch Distributed Algorithms 6.046J, Spring, 205 Nancy Lynch What are Distributed Algorithms? Algorithms that run on networked processors, or on multiprocessors that share memory. They solve many kinds of problems:

More information

Informed search algorithms

Informed search algorithms Informed search algorithms This lecture topic Chapter 3.5-3.7 Next lecture topic Chapter 4.1-4.2 (Please read lecture topic material before and after each lecture on that topic) Outline Review limitations

More information

CSC384: Intro to Artificial Intelligence Search III

CSC384: Intro to Artificial Intelligence Search III CSC384: Intro to Artificial Intelligence Search III 4 Prolog Tutorials: T1, T2, T3, and T4 T1: Fri Sep 22 10-11am SS2127 T2: Fri Sep 22 4-5pm BA3012 T3: Fri Sep 29 10-11am SS2127 T4: Fri Sep 29 4-5pm BA3012

More information

Problem Solving & Heuristic Search

Problem Solving & Heuristic Search 190.08 Artificial 2016-Spring Problem Solving & Heuristic Search Byoung-Tak Zhang School of Computer Science and Engineering Seoul National University 190.08 Artificial (2016-Spring) http://www.cs.duke.edu/courses/fall08/cps270/

More information

Informed search algorithms. (Based on slides by Oren Etzioni, Stuart Russell)

Informed search algorithms. (Based on slides by Oren Etzioni, Stuart Russell) Informed search algorithms (Based on slides by Oren Etzioni, Stuart Russell) Outline Greedy best-first search A * search Heuristics Local search algorithms Hill-climbing search Simulated annealing search

More information

More Planning and Prolog Operators

More Planning and Prolog Operators More Planning and Prolog Operators Artificial Intelligence Programming in Prolog Lecturer: Tim Smith Lecture 16 22/11/04 22/11/04 AIPP Lecture 16: More Planning and Operators 1 Planning continued Contents

More information

Applying Search Based Probabilistic Inference Algorithms to Probabilistic Conformant Planning: Preliminary Results

Applying Search Based Probabilistic Inference Algorithms to Probabilistic Conformant Planning: Preliminary Results Applying Search Based Probabilistic Inference Algorithms to Probabilistic Conformant Planning: Preliminary Results Junkyu Lee *, Radu Marinescu ** and Rina Dechter * * University of California, Irvine

More information

Path Planning. CMPS 146, Fall Josh McCoy

Path Planning. CMPS 146, Fall Josh McCoy Path Planning CMPS 46, Fall 203 Josh McCoy Readings Reading Path Planning: 97-255 Dijkstra and graphs early. CMPS 46, Fall 203 Pathfinding Why search? Consider this problem given a set of integers, we

More information

CPS 170: Artificial Intelligence Search

CPS 170: Artificial Intelligence   Search CPS 170: Artificial Intelligence http://www.cs.duke.edu/courses/spring09/cps170/ Search Instructor: Vincent Conitzer Search We have some actions that can change the state of the world Change resulting

More information

Branch & Bound (B&B) and Constraint Satisfaction Problems (CSPs)

Branch & Bound (B&B) and Constraint Satisfaction Problems (CSPs) Branch & Bound (B&B) and Constraint Satisfaction Problems (CSPs) Alan Mackworth UBC CS 322 CSP 1 January 25, 2013 P&M textbook 3.7.4 & 4.0-4.2 Lecture Overview Recap Branch & Bound Wrap up of search module

More information

Searching. Assume goal- or utilitybased. Next task to achieve is to determine the best path to the goal

Searching. Assume goal- or utilitybased. Next task to achieve is to determine the best path to the goal Searching Assume goal- or utilitybased agents: state information ability to perform actions goals to achieve Next task to achieve is to determine the best path to the goal CSC384 Lecture Slides Steve Engels,

More information

CS 331: Artificial Intelligence Informed Search. Informed Search

CS 331: Artificial Intelligence Informed Search. Informed Search CS 331: Artificial Intelligence Informed Search 1 Informed Search How can we make search smarter? Use problem-specific knowledge beyond the definition of the problem itself Specifically, incorporate knowledge

More information

Outline. Informed Search. Recall: Uninformed Search. An Idea. Heuristics Informed search techniques More on heuristics Iterative improvement

Outline. Informed Search. Recall: Uninformed Search. An Idea. Heuristics Informed search techniques More on heuristics Iterative improvement Outline Informed Search ECE457 Applied Artificial Intelligence Fall 2007 Lecture #3 Heuristics Informed search techniques More on heuristics Iterative improvement Russell & Norvig, chapter 4 Skip Genetic

More information

Silvia Rossi! Planning. Lezione n.! Corso di Laurea:! Informatica! Insegnamento:! Sistemi! multi-agente! ! A.A !

Silvia Rossi! Planning. Lezione n.! Corso di Laurea:! Informatica! Insegnamento:! Sistemi! multi-agente!  ! A.A ! Silvia Rossi! Planning 4 Lezione n.! Corso di Laurea:! Informatica! Insegnamento:! Sistemi! multi-agente! Email:! silrossi@unina.it! A.A. 2014-2015! Agent Planning! (R&N:11.1,11.2) Additional reading:

More information

CS 331: Artificial Intelligence Informed Search. Informed Search

CS 331: Artificial Intelligence Informed Search. Informed Search CS 331: Artificial Intelligence Informed Search 1 Informed Search How can we make search smarter? Use problem-specific knowledge beyond the definition of the problem itself Specifically, incorporate knowledge

More information

Route planning / Search Movement Group behavior Decision making

Route planning / Search Movement Group behavior Decision making Game AI Where is the AI Route planning / Search Movement Group behavior Decision making General Search Algorithm Design Keep a pair of set of states: One, the set of states to explore, called the open

More information

Informed search strategies (Section ) Source: Fotolia

Informed search strategies (Section ) Source: Fotolia Informed search strategies (Section 3.5-3.6) Source: Fotolia Review: Tree search Initialize the frontier using the starting state While the frontier is not empty Choose a frontier node to expand according

More information

6.825 Midterm Exam Fall 2002

6.825 Midterm Exam Fall 2002 6.825 Midterm Exam Fall 2002 October 29, 2002 1. (8 points) Convert the following four sentences to clausal form: (a) x. y.p(x) r(x, y) p(x 1 ) r(x 2, F (x 2 )) (b) x. y.p(x) r(x, y) p(a) r(a, y 1 ) (c)

More information

Informed search algorithms. Chapter 4

Informed search algorithms. Chapter 4 Informed search algorithms Chapter 4 Material Chapter 4 Section 1 - Exclude memory-bounded heuristic search 3 Outline Best-first search Greedy best-first search A * search Heuristics Local search algorithms

More information

Implicit Abstraction Heuristics for Cost-Optimal Planning

Implicit Abstraction Heuristics for Cost-Optimal Planning Abstraction for Cost-Optimal Planning Michael Katz Advisor: Carmel Domshlak Technion - Israel Institute of Technology Planning Language Classical Planning in SAS + Planning task is 5-tuple V, A, C, s 0,

More information

Uninformed Search Methods

Uninformed Search Methods Uninformed Search Methods Search Algorithms Uninformed Blind search Breadth-first uniform first depth-first Iterative deepening depth-first Bidirectional Branch and Bound Informed Heuristic search Greedy

More information

EECS 219C: Computer-Aided Verification Boolean Satisfiability Solving. Sanjit A. Seshia EECS, UC Berkeley

EECS 219C: Computer-Aided Verification Boolean Satisfiability Solving. Sanjit A. Seshia EECS, UC Berkeley EECS 219C: Computer-Aided Verification Boolean Satisfiability Solving Sanjit A. Seshia EECS, UC Berkeley Project Proposals Due Friday, February 13 on bcourses Will discuss project topics on Monday Instructions

More information

CS 416, Artificial Intelligence Midterm Examination Fall 2004

CS 416, Artificial Intelligence Midterm Examination Fall 2004 CS 416, Artificial Intelligence Midterm Examination Fall 2004 Name: This is a closed book, closed note exam. All questions and subquestions are equally weighted. Introductory Material 1) True or False:

More information

Classical Planning Problems: Representation Languages

Classical Planning Problems: Representation Languages jonas.kvarnstrom@liu.se 2017 Classical Planning Problems: Representation Languages History: 1959 3 The language of Artificial Intelligence was/is logic First-order, second-order, modal, 1959: General

More information

Module 4. Constraint satisfaction problems. Version 2 CSE IIT, Kharagpur

Module 4. Constraint satisfaction problems. Version 2 CSE IIT, Kharagpur Module 4 Constraint satisfaction problems Lesson 10 Constraint satisfaction problems - II 4.5 Variable and Value Ordering A search algorithm for constraint satisfaction requires the order in which variables

More information

CS 4700: Artificial Intelligence

CS 4700: Artificial Intelligence CS 4700: Foundations of Artificial Intelligence Fall 2017 Instructor: Prof. Haym Hirsh Lecture 7 Extra Credit Opportunity: Lecture Today 4:15pm Gates G01 Learning to See Without a Teacher Phillip Isola

More information

Outline. Best-first search

Outline. Best-first search Outline Best-first search Greedy best-first search A* search Heuristics Local search algorithms Hill-climbing search Beam search Simulated annealing search Genetic algorithms Constraint Satisfaction Problems

More information

Downloaded from ioenotes.edu.np

Downloaded from ioenotes.edu.np Chapter- 3: Searching - Searching the process finding the required states or nodes. - Searching is to be performed through the state space. - Search process is carried out by constructing a search tree.

More information