Application of Artificial Intelligence Techniques to Digital Computer Control of Nuclear Reactors

This paper investigates the application of two artificial intelligence techniques, heuristic programming and learning theory, to the problem of digital computer control of nuclear reactors. The purpose of using an artificial intelligence approach is to alleviate the requirement for a detailed mathematical model of the nuclear reactor system. A control system is developed ivhich demonstrates the primary features of such an approach. Results are shown for several reactor and control problem variations using a computer simulated reactor.