An intelligent user support for the two phases of simulation execution was newly developed for Module-based Simulation System(MSS). The MSS has been in development as a flexible simulation enviroment to improve software productivity in complex, large-scale dynamic simulation of nuclear power plant.
The AI programing by Smalltalk-80 was applied to materialize the two user-interface programs for (i) semantic diagnosis of the simulation program generated automarically by MSS, and (ii) consulation system by which user can set up consistent numarical input data files necessary for executing a MSS-generated progeam. Frame theory was utilized in those interface programs to represent the four knowledge bases, which are (i) usage information on module library in MSS and MSS-generated program, and (II) expertise knowledge on nuclear power plant analysis such as material properties and reactor system configuration.
Capabilities of those interface programs were confirmed by some example practice on LMFBR reactor dynamic calculation, and it was demonstrated that the knowledege-based systemization was effective to improve software work enviroment.