18-planning-t - Foundations of Artificial Intelligence...

Info iconThis preview shows pages 1–3. Sign up to view the full content.

View Full Document Right Arrow Icon

Info iconThis preview has intentionally blurred sections. Sign up to view the full version.

View Full DocumentRight Arrow Icon
This is the end of the preview. Sign up to access the rest of the document.

Unformatted text preview: Foundations of Artificial Intelligence Planning CS472 – Fall 2007 Thorsten Joachims Planning A planning agent will construct plans to achieve its goals, and then execute them. Analyze a situation in which it finds itself and develop a strategy for achieving the agent’s goal. Achieving a goal requires finding a sequence of actions that can be expected to have the desired outcome. Problem Solving Representation of actions actions generate successor states Representation of states all state representations are complete Representation of goals contained in goal test and heuristic function Representation of plans unbroken sequence of actions leading from initial to goal state Planning Example GOAL: Get a quart of milk and a bunch of bananas and a variable-speed cord-less drill. Planning vs. Problem Solving 1. Open up the representation of states, goals and actions. • States and goals represented by sets of sentences – Have ( Milk ) • Actions represented by rules that represent their preconditions and effects: Buy ( x ) achieves Have ( x ) and leaves everything else unchanged This allows the planner to make direct connections between states and actions. Planning vs. Problem Solving 2. Most parts of the world are independent of most other parts. • Can solve using divide-and-conquer strategy. • Can re-use sub-plans (go to supermarket) Planning vs. Problem Solving 3. Planner is free to add actions to the plan wherever they are needed, rather than in an incremental sequence starting at the initial state. • No connection between the order of planning and the order of execution. • Representation of states as sets of logical sentences makes this freedom possible. Planning as a Logical Inference Problem Axioms: On(A,C) On(C,Table), On(D,B), On(B,Table), Clear(A), Clear(D) Plus rules for moving things around… Prove: On (A,B) ® On(B,C) Planning as Deduction: Situation Calculus In first-order logic, once a statement is shown to be true, it remains true forever. Situation calculus: way to describe change in first-order logic. Situation Calculus Fluents: functions and predicates that vary from one situation to the next on ( A,C ) on ( A,C,S ) at ( agent ,[1,1]) at ( agent , [1,1], S ) Atemporal functions and predicates: true in any situation block (A) gold (G 1 ) Situation Calculus: Actions Actions are described by stating their effects. Possibility Axiom: preconditions Poss(a,s ). Effect Axiom: Poss(a,s) Changes that result from action....
View Full Document

This note was uploaded on 02/19/2008 for the course CS 4700 taught by Professor Joachims during the Fall '07 term at Cornell.

Page1 / 8

18-planning-t - Foundations of Artificial Intelligence...

This preview shows document pages 1 - 3. Sign up to view the full document.

View Full Document Right Arrow Icon
Ask a homework question - tutors are online