ANN-based real-time parameter optimization via GA for superheater model in power plant simulator

In order to rapidly optimize superheater model parameters to achieve the required precision, ANN and GA are combined to solve the problem. Since the classic optimization methods are not appropriate for mechanism model in power plant simulator, GA is applied to optimize model parameters. Input data, output data of model and optimized parameters are normalized to make learning sample. After ANN is trained with back-propagation algorithm, it is able to optimize model parameters in real-time. Simulation result shows that superheater model optimized by this method achieves the required accuracy. The method replaces manual parameter regulation and shortens optimization time. It is a general method, provides a new way for parameter optimization for thermal equipment model in power plant simulator.