Heuristic Search Planners
Heuristic search planners search state space with a heuristic that can be extracted automatically from the problem encoding, and then combined with standard search algorithms. Success is influenced by factors such as the choice of heuristic, search algorithm and search direction. They are arranged here in alphabetical order.
AltAlt from Rao Kambhampati's Yochan research group at Arizona State University.
- Description by author: AltAlt started out as "a little of this,
and a little of that". It is a student-only effort where a group of graduate
students parallely toyed with a number of alternative planning ideas and
evaluated them before fielding the better one. The actual idea used in the
competition, GP-HSP, which became the official AltAlt planner, uses effective
and admissible heuristics extracted from the planning graph to drive the
backward state space search. AltAlt was developed in Lisp and later ported
to C for the competition (round1). A problem we faced was that the relatively
newer C version was hard to try better heuristics with, as the competition
progressed, even when newer heuristics were easily seen to be feasible in the
Lisp version.
- Publications
- Source Code: Remote, Local
- Platform: Red Hat Linux 6.0
- Input Language:
PDDL
- Implementation Language: C & Lisp
- See Also: Graph-based planners
FF and Metric-FF from Jörg Hoffmann at Albert-Ludwigs University, Freiburg. Awarded for Outstanding Performance at the 2nd International Planning Competition and Top Performer in the Strips Track of the 3rd International Planning Competition.
- Description by author: FF (Fast-Forward) is basically a variation
of HSP. Interpreting the HSP heuristic as a special case of the graph building
phase in GRAPHLAN, it extends the heuristic to also employ GRAPHPLAN's plan
extraction phase. The length of the plans that GRAPHPLAN returns give a
heuristic goal distance measure, and the actions that GRAPHPLAN selects can
be used for guiding search.
Using these informations, FF employs a novel local search strategy
that combines Hill-climbing with systematic search. It makes use of the fact
that phenomena like big plateaus or local minima — with respect to the
heuristic described above — do not occur very often in benchmark
planning problems.
Metric-FF is an extension of the FF planner to
(ADL combined with) numerical state variables, more precisely to PDDL 2.1
level 2, yet more precisely to the subset of PDDL 2.1 level 2 with the
algorithmic principles.
-
Publications:
- J. Hoffmann, B. Nebel, The
FF Planning System: Fast Plan Generation Through Heuristic Search,
in: Journal of Artificial Intelligence Research, Volume 14, 2001, Pages
253 - 302.
- J. Hoffmann, FF: The
Fast-Forward Planning System, in: AI Magazine, Volume 22, Number 3,
2001, Pages 57 - 62.
- J. Hoffmann, The Metric-FF Planning System: Translating ``Ignoring Delete Lists'' to Numerical State Variables, submitted to: Journal of Artificial Intelligence Research, special issue on the 3rd
International Planning Competition.
Source Code: - FF v-1.0(Strips): Remote,
Local
- FFv-2.3(ADL):
Remote, Local
- Metric-FF: Remote,
Local
- Input Language: PDDL, PDDL2.1 &ADL
- Implementation Language: C
- See Also: Graph-based planners
GRT, MO-GRT from Ioannis Refanidis at University of Macedonia, Thessaloniki, Greece.
- Description by author:GRT is a domain-independent heuristic
planner, which uses pure STRIPS representation and forward search.
Characterised by the production of a Greedy Regression Table-based
heuristic during a pre-processing phase. A recent extension of GRT,
named MO-GRT, constructs and evaluates plans on the basis
of multiple criteria.
- Publications
- Source Code: Remote, Local
- Implementation Language: C++
HSP and HSP2.0 from Hector Geffner and Blai Bonet,
Universidad Simon Bolivar.
- Description by author:HSP is a planner based on the ideas
of heuristic search.1Heuristic search algorithms perform forward search
from an initial state to a goal state using an heuristic function that
provides an estimate of the distance to the goal.
HSP2.0 descends from the HSP planner. The idea
is similar: planning instances are mapped into state-space search problems
that are solved with heuristics extracted from the representation.
Unlike the original HSP planner, the new version supports forward and
backward search, several heuristics, and a more expressive modeling
language (a fragment of typed ADL). The search algorithm is Weighted A*;
an A* search with the heuristic multiplied by a constant W > 1.
There are also a number of improvements in the code.
Different options (search direction/heuristic) can be run in sequence or
concurrently.
- Publications
- Source Code: Remote,
Local: HSP-2.0
- Input Language: PDDL and ADL
- Implementation Language: C
LPG developed by Alfonso Gerevini and Ivan Serina from the University of Brescia.
SAPA from Minh B. Do and Rao Kambhampati from Yochan research group at Arizona State University.
VHPOP from Håkan Younes at Carnegie Mellon University.
- Description by author:VHPOP is a versatile heuristic partial order planner, loosely based on UCPOP, with the following main features:
- Can plan with either ground or lifted actions.
- Can enforce joint parameter domain constraints when planning with lifted actions.
- Has several different plan ranking heuristics to choose from that can be combined into complex plan ranking functions.
- Efficiently implements all common flaw selection strategies, including LCFR and ZLIFO, and many novel ones, and allows flaw selection strategies to be specified using a concise notation.
- Several flaw selection strategies can be used simultaneously.
- Implements A*, IDA*, and hill climbing search.
- Fully supports levels 1 and 3 of PDDL+.
Publications:
- Håkan L. S. Younes and Reid G. Simmons.
On the role of ground actions in refinement planning.
In Malik Ghallab, Joachim Hertzberg, and Paolo Traverso, editors, Proceedings of the Sixth International Conference on Artificial Intelligence Planning and Scheduling Systems, pages 54-61, Toulouse, France, April 2002.
AAAI Press.
- Source Code: Remote,
Local
- Platform: RedHat Linux5.2, 6.2, 7.0, 7.1, IRIX 6.5 and SunOS 5.7
- Input Language: PDDL
- Implementation Language: C++
 |
|  |
|