Temporal reasoning in logic programming: a case for the situation calculus
Javier A. Pinto, Raymond Reiter · 1993
We propose, and axiomatize, an extended version of the situation calculus [10] for temporal reasoning in a logic programming framework. This extended language provides for a linear temporal structure, which may be viewed as a path of actual event occurrences within the tree of possible situations of the "classical" situation calculus. The extended language provides for events to occur and fluents to hold at specific points in time. As a result, it is possible to establish a close correspondence between this extended situation calculus and other linear time formalisms which have been proposed in opposition to the situation calculus. In particular, we argue that the functionality of the event calculus [6] is subsumed by the extended situation calculus. We present a logic program for temporal reasoning which is provably sound for our axiomatization, relative to the Clark completion semantics of the program. Our logic programming approach has the advantage of being grounded in a pure (without negation as failure) first order axiomatization suitable for reasoning about events and their occurrences. Moreover, efficient algorithms can be obtained for a suitable class of temporal reasoning problems, following the ideas of Kowalski [5].