Modeling the Temporal Behavior of Technical Systems
Many inference-mechanisms that draw conclusions from given facts or measurements using physical relations are based on the propagation of values by constraints. Such inference-mechanisms are often limited in their ability to consider temporal relations. However, the constraint idea also provides a framework for reasoning about temporal behavior. In [Williams86], TCP (Temporal Constraint Propagator), a method for integrating time into constraint systems is presented. In contrast to simple constraint systems, propagated objects are not values but pairs consisting of a value and a time interval, called episodes. In our paper we present another system for propagating episodes, EP (Episode Propagator), that overcomes some limitations of Williams TCP. Because an episode contains information about when a variable adopts a value, temporal behavior can be modeled by propagating sets of episodes using EP. This paper emphasizes the use of EP in modeling digital and synchronous circuits.
Unable to display preview. Download preview PDF.
- [Allen83]Allen, J.: Maintaining Knowledge about Temporal Intervals. Communications of the ACM. pp. 832–843. November 1983Google Scholar
- [deKleer, Brown84]deKleer, J.; Brown, J.S.: A qualitative Physics based on Confluences. AI Journal 24. pp. 7–83. 1984Google Scholar
- [Stallman, Sussman77]
- [Steele80]Steele, G.L.: The Definition and Implementation of a Computer Language based on Constraints. MIT AI–TR–595. 1980Google Scholar
- [Sussman, Abelson85]Sussman, G.J.; Abelson, H.: Structure and Interpretation of Computer Programs. MIT Press. 1985Google Scholar
- [Sussman, Steele80]Sussman, G.J.; Steele, G.L.: Constraints – A Language for expressing almost hierarchical descriptions. AI Journal 14. pp. 1–40. 1980Google Scholar
- [Williams86]Williams, B.: Doing Time: Putting Qualitative Reasoning on firmer Ground. AAAI–86 Proceedings, pp. 105–112. 1986Google Scholar