Use of an external optimizing algorithm with a GPSS model

  • Authors:
  • Robert M. Lefkowits;Thomas J. Schriber

  • Affiliations:
  • -;-

  • Venue:
  • WSC '71 Proceedings of the 5th conference on Winter simulation
  • Year:
  • 1971

Quantified Score

Hi-index 0.00

Visualization

Abstract

Although GPSS enjoys many advantages as a language for modeling discrete, non-deterministic systems, it suffers from several disadvantages. GPSS is interpretive, resulting in relatively slow execution times. In addition, the input/output and computational capabilities in the language are limited. Disadvantages like these can be lessened or overcome by use of the GPSS HELP Block, which makes it possible to interface an executing GPSS model with one or more FORTRAN subroutines. As an instructive example of HELP Block utility, this paper shows how a GPSS model has been mated with a FORTRAN subroutine implementing the univariate searching strategy. The result is a GPSS-FORTRAN combination which automates the search for the optimal way to configure a system.