甲醇精馏塔温压解耦控制系统的研究  

Study on temperature and pressure decoupling control system for methanol distillation column

作　　者：龚瑞昆 张文庆 龚雨含 周国庆 GONG Ruikun;ZHANG Wenqing;GONG Yuhan;ZHOU Guoqing(North China University of Science and Technology,Tangshan 063000,China;Tangshan Polytechnic College,Tangshan 063000,China)

机构地区：[1]华北理工大学,河北唐山063000 [2]唐山工业职业技术学院,河北唐山063000

出　　处：《现代电子技术》2021年第23期83-86,共4页Modern Electronics Technique

基　　金：国家自然科学基金(61203343)。

摘　　要：甲醇精馏是一个复杂的工艺过程,具有多变量、非线性和强耦合等特点,目前大多采用传统的PID控制,然而控制变量塔顶温度与塔顶的大气压强之间耦合严重,控制精度不高,从而影响产品的质量。因此文中对塔顶温度与塔顶大气压强之间的数量关系建立解耦模型,采用RBF前馈人工神经网络算法,并结合FPGA强大的并行处理能力,真正实现人工神经网络的快速性。与传统的PID控制进行对比,文中模型表现出调节时间更短、超调量更小和上升时间更短等性能优势,从而使得甲醇精馏塔的精馏过程更优化、产品质量更好、经济效益更高,具有一定的实际意义。通过这一研究也使得综合的性能表现向着更为高质量的方向发展。Methanol distillation is a complex process,which has the characteristics of multivariable,nonlinear and strong coupling.At present,the traditional PID control method is mostly adopted.However,the coupling between the temperature of the tower top and the atmospheric pressure of the tower top is serious,and the control accuracy is not high,which affects the quality of products.Therefore,a decoupling model is established for the quantitative relationship between the temperature of the tower top and the atmospheric pressure of the tower top.In the model,the RBF(radial basis function)feed forward artificial neural network algorithm is adopted,and the rapidity of the artificial neural network is truly realized in combination with the powerful parallel processing ability of FPGA(field programmable gate array).In comparison with the traditional PID control,this model shows the performance advantages of shorter adjustment time,smaller overshoot and shorter rise time,so that the distillation process of methanol distillation column is more optimized,the product quality is better and the economic benefit is higher,which has a certain practical significance.This research makes the comprehensive performance of the system developed towards higher quality.

关 键 词：温压解耦控制 甲醇精馏塔 人工神经网络 解耦模型 RBF FPGA 

分 类 号：TN98-34[电子电信—信息与通信工程] TP11[自动化与计算机技术—控制理论与控制工程]