Efficient Temporal Planning Using Metastates

Abstract

When performing temporal planning as forward state-space search, effective state memoisation is challenging. Whereas in classical planning, two states are equal if they have the same facts and variable values, in temporal planning this is not the case: as the plans that led to the two states are subject to temporal constraints, one might be extendable into at temporally valid plan, while the other might not. In this paper, we present an approach for reducing the state space explosion that arises due to having to keep many copies of the same ‘classically’ equal state – states that are classically equal are aggregated into metastates, and these are separated lazily only in the case of temporal inconsistency. Our evaluation shows that this approach, implemented in OPTIC and compared to existing state-of-the-art memoisation techniques, improves performance across a range of temporal domains.

Higher-Dimensional Potential Heuristics: Lower Bound Criterion and Connection to Correlation Complexity

“…2 Our definition is similar to Coles, Fox, and Smith (2007) but in a finite domain representation instead of a STRIPS like representation, and they allow combining operators only if they applicable in succession, where we allow arbitrary combinations but ensure that the resulting macro is not applicable if the operators are not applicable in succession.…”

Automated Machine Learning Service Composition

“…On top of this formalism, we build a hierarchical model [22]. A task network is a partially ordered set T of tasks.…”