Building planning systems that operate in real domains requires coping with both uncertainty and time pressure. This paper describes a model of reaction plans, which are generated using a formalization of actions and of state descriptions in probabilistic logic, as a basis for anytime planning under uncertainty. The model has the following main features. At the action level, we handle incomplete and ambiguous domain information, and reason about alternative action effects whose probabilities are given. On this basis, we generate reaction plans that specify different courses of action, reflecting the domain uncertainty and alternative action effects; if generation time was insufficient, these plans may be left unfinished, but they can be reused, incrementally improved, and finished later. At the planning level, we develop a framework for measuring the quality of plans that takes domain uncertainty and probabilistic information into account using Markov chain theory; based on this framework, one can design anytime algorithms focusing on those parts of an unfinished plan first, whose completion promises the most "gain". Finally, the plan quality can be updated during execution, according to additional information acquired, and can therefore be used for on-line planning. 1
|
481
|
An analysis of time-dependent planning
– Dean, Boddy
- 1988
|
|
466
|
J.L.: Finite Markov Chains
– Kemeny, Snell
- 1960
|
|
353
|
Dynamic Programming and Markov Processes
– Howard
- 1960
|
|
348
|
Learning and executing generalized robot plans
– Fikes, Hart, et al.
- 1972
|
|
315
|
Universal plans for reactive robots in unpredictable environments
– Schoppers
- 1987
|
|
314
|
Probabilistic logic
– Nilsson
|
|
312
|
A model for reasoning about persistence and causation
– Dean, Kanazawa
- 1989
|
|
304
|
On the difference between updating a knowledge base and revising it, in P. Gärdenfors ed.: Belief Revision
– KATSUNO
- 1992
|
|
291
|
Planning and Control
– Dean, Wellman
- 1991
|
|
239
|
Practical Planning: Extending the Classical AI Planning Paradigm
– Wilkins
|
|
239
|
Reasoning about action using a possible models approach
– Winslett
- 1988
|
|
236
|
Knowledge in Flux
– GÄRDENFORS
- 1988
|
|
210
|
Solving TimeDependent Planning Problems
– Boddy, Dean
- 1989
|
|
201
|
Conditional nonlinear planning
– Peot, Smith
- 1992
|
|
145
|
Planning under time constraints in stochastic domains
– Dean, Kaelbling, et al.
- 1995
|
|
122
|
Reasoning about action I: A possible worlds approach
– Ginsberg, Smith
- 1988
|
|
109
|
Anytime synthetic projection: Maximizing the probability of goal satisfaction
– Drummond, Bresina
- 1990
|
|
97
|
Achieving several goals simultaneously
– Waldinger
- 1977
|
|
92
|
Situated control rules
– Drummond
- 1989
|
|
65
|
How to do things with worlds: on formalizing actions and plans
– Brewka, Hertzberg
- 1993
|
|
60
|
A Model-Based Belief Revision System
– Chou, Winslett
- 1994
|
|
60
|
Generating Conditional Plans and Programs
– Warren
- 1976
|
|
58
|
Representations for decision-theoretic planning: Utility functions for deadline goals
– Haddawy, Hanks
- 1992
|
|
56
|
Probabilities of Conditionals and Conditional Probabilities
– Lewis
- 1976
|
|
51
|
Practical temporal projection
– Hanks
- 1990
|
|
48
|
Decision theory and artificial intelligence ii: The hungymonkey. Cognitive Science 1:158-192
– Feldman, Sproull
- 1977
|
|
45
|
Modular utility representation for decisiontheoretic planning
– Wellman, Doyle
- 1992
|
|
41
|
Toward Agent Programs with Circuit Semantics
– Nilsson
- 1991
|
|
40
|
A method of computing generalized Bayesian probability values for expert systems
– Cheeseman
- 1983
|
|
38
|
Issues in decision-theoretic planning: Symbolic goals and numeric utilities
– Haddawy, Hanks
- 1990
|
|
33
|
Sensible planning: Focusing perceptual attention
– Chrisman, Simmons
- 1991
|
|
28
|
Declarative Goals in Reactive Plans
– Beetz, McDermott
- 1992
|
|
26
|
A model for projection and action
– Kanazawa, Dean
- 1989
|
|
24
|
Reaction-first search
– Drummond, Swanson, et al.
- 1993
|
|
24
|
On optimal gametree search using rational metareasoning
– Russell, Wefald
- 1989
|
|
20
|
Synthesis of Reactive Plans for Multi-Path Environments
– Kabanza
- 1990
|
|
19
|
Optimal probabilistic and decision-theoretic planning using markovian decision theory
– Koenig
- 1992
|
|
18
|
Computing Knowledge Base Updates
– Val
- 1992
|
|
15
|
A representation of action and belief for automatic planning systems
– Drummond
- 1986
|
|
15
|
A semi-reactive planner based on a possible models action formalization
– Thi'ebaux, Hertzberg
- 1992
|
|
14
|
A Temporal Revision Model for Reasoning about World Change
– CORDIER, SIEGEL
- 1992
|
|
13
|
defense of reaction plans as caches
– In
- 1989
|
|
9
|
Belief revisions and updates in numerical formalisms. an overview, with new results for the possibilistic framework
– Dubois, Prade
- 1993
|
|
9
|
TCA*--A Time-Constrained Approximate A* Search Algorithm
– Wah, Chu
- 1990
|
|
3
|
R'evision des croyances et r'evision du raisonnement pour la planification. PhD thesis, Ecole Nationale Sup'erieure de l'A'eronautique et de l'Espace
– Garcia
- 1993
|
|
2
|
Artificial Intelligence Planning Systems
– Hendler, editor
- 1992
|
|
2
|
Anytime Reaction Planning in Probabilistic Logic
– Thi'ebaux
- 1992
|
|
2
|
Efficient ressource-bounded reasoning in AT-RALPH
– Zilberstein, Russel
- 1992
|
|
1
|
qwertz. a toolbox for building AI planners
– Gordon, Hertzberg, et al.
|
|
1
|
Reasoning under varying und uncertain resource constraints
– Horovitz
- 1988
|
|
