神经网络方法在循环流化床燃烧中的应用

The application of the neural network control method in the combustion process of circulating fluidized bed boiler

本文分析了循环流化床锅炉燃烧控制系统的特点。针对主蒸汽压力控制系统,提出一种新型的基于改进BP神经网络的PID控制方法,该方法将神经网络和PID控制策略相结合,既具有神经网络自学习、自适应及逼近任意函数的能力,又具有常规PID控制器结构简单的特点。仿真研究表明这种BP神经网络PID控制能克服对象的大惯性、抗干扰性,大大改善控制品质。

The article analyses the feature of the combustion control system of circulating fluidized bed boiler. To the control system of the main steam pressure, this paper proposes a new type of PID control method based on the improved BP neural network, which is a combina- tion of the neural network and the PID strategy, It has the merits of both neural network and PID controller. A simulation study showed that the platform is good in overcoming the object inertia, time--variability, and random disturbance ,and can considerably improve the control quality.