Heuristic Search for Multi-Objective Probabilistic Planning

TL;DR

This paper extends heuristic search methods to solve multi-objective probabilistic planning problems, specifically multi-objective stochastic shortest paths, by designing new algorithms and heuristics that improve policy computation.

Contribution

It introduces MOLAO* and MOLRTDP algorithms for multi-objective stochastic shortest path problems, expanding heuristic search to more expressive probabilistic planning scenarios.

Findings

Algorithms outperform existing methods in complex domains

Heuristics effectively guide search in multi-objective stochastic settings

Coverage sets of non-dominated policies are efficiently computed

Abstract

Heuristic search is a powerful approach that has successfully been applied to a broad class of planning problems, including classical planning, multi-objective planning, and probabilistic planning modelled as a stochastic shortest path (SSP) problem. Here, we extend the reach of heuristic search to a more expressive class of problems, namely multi-objective stochastic shortest paths (MOSSPs), which require computing a coverage set of non-dominated policies. We design new heuristic search algorithms MOLAO* and MOLRTDP, which extend well-known SSP algorithms to the multi-objective case. We further construct a spectrum of domain-independent heuristic functions differing in their ability to take into account the stochastic and multi-objective features of the problem to guide the search. Our experiments demonstrate the benefits of these algorithms and the relative merits of the heuristics.