BP神经网络模糊PID控制在加热炉煤气流量控制中的应用

钢厂中加热炉是一个复杂的受控对象,存在着非线性、时变性、纯滞后因素和不确定随机干扰等因素.从对其燃烧状况的分析来看,加热炉温度的调节主要是靠对煤气流量的控制来完成的,因而确立一种合理的煤气流量控制方案是实现加热炉燃烧智能化控制的关键.本文提出将模糊控制、PID控制和神经网络三种技术相结合,应用于煤气流量进行控制.仿真研究表明这种BP神经网络模糊PID控制在克服对象的大惯性、抗干扰性、非线性和纯滞后上,大大改善控制品质.

Application of fuzzy PID control based on BP neural network in coal gas flux control for heating furnace

Being a sophisticated object, the heating furnace employed in a steel-making company usually encounters in its control such problems as the non-linearity, time-variability, pure lag and random disturbance. According to an analysis of the burning condition of a furnace, the temperature of the furnace can be controlled mainly by controlling the coal gas flux. Therefore, a key point in realizing an intelligent control of the furnace coal gas flux would be to establish a proper platform for the control. A control platform combining fuzzy control, PID control and BP neural network was then proposed and applied to the coal gas flux control. A simulation study showed that the platform is good in overcoming the object inertia, time-variability, pure lag and random disturbance, and can thus considerably improve the control quality.