锅炉过热器系统机理与神经网络组合建模方法

A PHYSICAL LAW AND NEURAL NETWORK INTEGRATED MODELING METHOD FOR BOILER SUPERHEATER SYSTEMS

电厂锅炉系统的模型多用集总参数法建模 ,文中通过分析指出了集总参数建模法方法本身所带来的不可弥补的缺陷 ,从而提出了高精度的机理与神经网络的组合建模方法 ,并结合过热器的建模过程对组合建模方法的应用作了阐述。此法对整个锅炉乃至电厂系统的精确建模也是适用的。

The boiler models in power plants are commonly constructed as lumped models, which are subject to certain inherent defects as shown in the analysis of this paper. Thus we propose a new precise method integrating the physical law and neural network modeling technique, and demonstrate the application of the method by the modeling of a superheater. The proposed method is also applicable to accurate modeling of such systems as power plant boiler and even the whole power plant.