神经网络模糊PID控制在加热炉煤气流量控制中的应用

The Application of Fuzzy PID Control Based on BP Neural Network to Flux Control of Coal Gas for the Furnace

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。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。

The furnace of steel company is a complex controlled object, and has the casual process of nonlinear and time-variability, pure lag and casual disturbance. Considering burning condition, the adjusting of temperature of furnace is accomplished mainly by coal gas flux control. Therefore, coal gas flux control is the key of realizing the intelligent control of the furnace. The thesis introduces the form of combining BP neural network and fuzzy system with PID control, and applies this method on coal gas flux control. The simulation research shows that the big inertia time-variability, pure lag and random disturbance can be overcome, and considerably improves control quality.