The aim of this project is to make the FMAP planning engine available in the unified_planning library by the AIPlan4EU project.
FMAP uses a distributed heuristic search strategy. Each planning agent in the platform features an embedded search engine based on a forward partial-order planning scheme FMAP leverages a forward-chaining partial-order planner (POP) that allows agents to plan their actions in parallel whenever possible, which largely improves the quality of the resulting solution plans. Moreover, the forward-chaining approach relies on the frontier state (state that results from executing the actions of a node) to compute accurate state-based estimates.
We recommend the installation from PyPi because it has pre-built wheels for all common operating systems.
To automatically get a version that works with your version of the unified planning framework, you can list it as a solver in the pip installation of unified_planning:
pip install unified-planning[fmap]
If you need several solvers, you can list them all within the brackets.
You can also install the FMAP integration separately (in case the current version of unified_planning does not include FMAP or you want to add it later to your unified planning installation). With
pip install up-fmap
you get the latest version. If you need an older version, you can install it with:
pip install up-fmap==<version number>
You can use fmap to solve a Multi-Agent problem. It allows the specification of a specific heuristic function used to evaluate the quality of the plans. The name of the custom parameter is heuristic and the following values are supported:
- 0 - FF heuristic: guides the search through the well-known h_FF heuristic function. This option is available for single-agent planning tasks only.
- 1 - DTG heuristic: evaluates plans via the heuristic h_DTG.
- 2 - default option - DTG + Landmarks: this option applies the multi-heuristic search scheme of the MH-FMAP solver by combining the h_DTG and h_Land heuristics to guide the search.
- 3 - Inc. DTG + Landmarks: incremental multi-heuristic mode that makes use of h_DTG and h_Land. You can for example call it as follows:
from unified_planning.shortcuts import *
from unified_planning.engines import PlanGenerationResultStatus
problem = MultiAgentProblem('myproblem')
# specify the problem (e.g. agents, agent public fluents, agent private fluents, environment fluents, initial state, agent actions, goal)
...
planner = OneshotPlanner(name="fmap")
result = planner.solve(problem)
if result.status == PlanGenerationResultStatus.SOLVED_SATISFICING:
print(f'{Found a plan.\nThe plan is: {result.plan}')
else:
print("No plan found.")
Notebooks:
Multi-Agent Plan Simple Example
Multi-agent planning
DTG + Landmarks: this option applies the multi-heuristic search scheme of the MH-FMAP solver (described in this paper) by combining the h_DTG and h_Land heuristics to guide the search.
Oneshot planning UP Documentation