摘要
精馏过程是化工、石油、医药等领域常见的生产过程,由于其非线性、时变性和强耦合等特点,使建模、控制与优化成为理论与实践研究中的难点。本文以一个二元丙烯精馏塔为研究对象,在由物料及能量平衡原理推导出精馏过程数学模型的基础之上,采用基于集结法的动力学模型降阶处理方法对原模型进行降阶,最后为该精馏塔模型设计了基于PID神经网络的解耦控制器。
Distillation process was the common production process in the chemical, petroleum, medicine and other fields.The modeling, control and optimization were the difficult problems in the theory and practice because of its nonlinear, time-varying and strong coupling characteristics. According to the material and energy balance principle, the modeling of a binary propylene distillation tower was derived, and was reduced order by using the aggregation method. Finally, the decoupling controller for distillation tower was designed based on PID-NN.
出处
《才智》
2013年第20期250-251,共2页
Ability and Wisdom
关键词
精馏
模型降阶
解耦
PID神经网络
distillation
model reduction
decoupling
PID neural network