Modeling of the Steam Pressure in Separator of a Once-through Unit Based on the ANN

It is challenging and interesting to establish a precise steam pressure in separator dynamic model of a once-through power plant unit in order to meet large scale load demands from the power grid and the needs of the engineering research. Steam pressure reveals the energy balance of boiler and turbine. This study proposes to establish such a steam pressure in separator mathematical model of a once-through boiler-turbine unit under dry operating conditions. Then the BP neural network model and the RBF neural network model were used to predict the steam pressure in separator, combined with running data. Moreover, in order to improve the performance of the network, the exhaustive method is used to optimize the parameters of the RBF network. After this, to further enhance model performance, the dynamic mathematical model was extended to be the one with different sets of parameters under different load ranges. More importantly, the dynamic model captures the essential dynamic characteristics of the unit. Therefore, the model can be feasible and applicable for simulation analysis and testing control algorithms. (paper)