Abstract
The subject of modeling and simulation is considered to be as one of the effective scientific tools for understanding system behaviors as well as testing and verifying of new control methods. In this paper. We modeled a nuclear power plant with sodium cooled fast reactor. The physical modeling considers; reactor core, heat exchangers, steam drums, preheaters, piping, pums, and valves. We simulated the plant through the control rod movements, as well as, cases of flow changes through valves opening and closure, and pumps being on and off. Most of the industrial complexes base their control methods on the well known PID - controllers. However, many of them are developing high interest in using modern control approaches. The nuclear industry, on the other hand, still cautious even with the availability of large theoretical studies which are performed by researchers that try to convince atomic authorities to transfer such studies which are performed by researchers that try to convince atomic authorities to transfer such studies to practical applications. In this work, we made study on the control of a nuclear reactor using neural network control. Such method has the advantage of not requiring a knowledge concise model of plant under investigation, it is the same advantage that is making PID- controllers quite famous and very suited for many control applications. Our complete work is developed in a computer package written in C-language. The package is developed with user friendly interface both in arabic and in english helps the user to browse freely through three phases, namely, preprocessor, simulation and control, and postprocessor for cross graphing analysis. Such package can be used as a tool for either a computer aide training CAT, or for computer aide learning CAL.