NN-based Gas Power Plant Simulator

M.R. Hesamzadeh, A. Salehi, A.R. Seifi, and A.A. Safavi (Iran)


Gas Power Plant, Simulator, SIMULINK, Neural Networks.


With the availability of powerful and high performance processors, advanced numerical methods, flexible and capable software, there is a great opportunity to develop high performance simulators for analysis of training on complex systems. Power plants are one group the complex systems with serious impact on economy, and operation of industries. A thorough understanding of power plants and their good operation and fast reaction to energy consumption is vital in any society. This paper addresses the development of a set of system component simulation modules combined with a control structure in a common software framework for a typical gas power plant. The simulator was designed for educational purposes using SIMULINK based on Object Oriented Programming (OOP) and C programming language. Since power plants are complicated systems and the simulation time is important, modeling approaches based on artificial neural networks are applied to the complex dynamics of the plant. This will lead to a faster and more flexible simulator, which can be easily matched to any real power plant of the same kind. The reduced simulation time also allows the utilization of the developed NN-based simulator for on-line control purposes. Results of these be presented for a typical 34MW power plant.

Important Links:

Go Back