蓄热式加热炉的神经网络燃烧控制

Neutral network combustion control for regenerative furnace

蓄热式加热炉温度对象具有非线性、大滞后的特点,本文运用神经网络的控制方法,在被控对象进行在线辨识的基础上,对神经网络权系数进行实时调整,使系统具有自学习、自适应性,仿真效果表明其控制效果优于一般PID控制。

The temperature control of regeneration furnace is a nonlinear and delaying object. This paper applies neutral network control method to adjust theweight coefficient of the neutral network based on system identification. The control system is of self-learning and self-adaptability. The simulation result shows that the control effect is preferable to PID.