An expert system framework based on a simulation generator

Abstract

Expert Systems (ES) implementations automatically perform tasks for which specially trained or talented people have been re quired. Fifth generation simulation systems integrate the tools developed in the fourth generation and capture the knowledge of the expert programmer as well as that of the simulation model ing expert. Haddock has programmed a user- oriented simula tion generator for the design and control of flexible manufac turing systems (FMS). The development of an ES based on the generator is described in this paper. The system to be described solely requires knowledge of the system to be simulated from the user. FORTRAN written sub routines, incorporated within the software structure of SIMAN, interpret the results of experimental runs and make statistical inferences about the performance measure. Simulation generators can assist simulationists in model develop ment and update, as well as in the analysis of alternative sce narios. A very desirable feature of Intelligent Front Ends (IFEs) is to have the capabilities of analyzing their output. These capa bilities not only reduce the total time required to perform the simulation, but also prevent its misuse.