Core Concepts
Metric Temporal Equilibrium Logic (MEL) provides a non-monotonic framework for specifying and reasoning about dynamic systems with both qualitative and quantitative temporal constraints.
Abstract
The paper introduces Metric Temporal Equilibrium Logic (MEL), a non-monotonic extension of Answer Set Programming (ASP) that allows for reasoning about dynamic systems with both qualitative and quantitative temporal constraints.
The authors start by defining the monotonic logic Metric Logic of Here-and-There (MHT), which extends the logic of Here-and-There with metric temporal operators. MHT allows for the representation of timed traces, where each state is associated with a specific time point.
The authors then define the non-monotonic MEL by selecting certain MHT models as equilibrium models. This provides a way to deal with inertia and other types of defaults in the representation of dynamic systems. The paper discusses several properties of MEL, including strong equivalence, which is shown to coincide with equivalence in the monotonic logic MHT.
Additionally, the authors provide a translation of MHT into a fragment of first-order logic called Quantified Here-and-There with Difference Constraints (QHT[≼δ]). This translation serves as a blueprint for implementing MEL using ASP modulo difference constraints.
The paper demonstrates the expressiveness of MEL through an example involving the behavior of traffic lights, where both qualitative and quantitative temporal constraints are specified.