焙烧烟气温度的建模与多变量解耦控制  

Modeling and Multivariable Decoupling Control for Anode Baking Flue Gas Temperature

作　　者：李晓斌[1] 吴燕翔[1] 寇得民[2] 

机构地区：[1]上海海洋大学工程学院,上海200090 [2]兰州理工大学电气与信息工程学院,甘肃兰州730050

出　　处：《控制工程》2009年第2期154-158,共5页Control Engineering of China

基　　金：国家科技攻关计划基金资助项目(2002BA901A28);甘肃省省长基金资助项目(GS015-A52-012)

摘　　要：阳极焙烧烟气温度是一个耦合的非线性的多变量控制对象,是提高电解铝生产质量和效率的主要参数,为实现其精确控制,依据现场实际测得的温度数据,辨识出烟气温度的控制模型。以此模型为初始预测模型,提出了一种多变量预测函数解耦控制方法。仿真研究了焙烧烟气温度解耦控制的有效性,与原有的焙烧烟气温度PID控制方法进行了实际应用比较。仿真和应用比较结果表明,这种控制方法的控制精度和鲁棒性优于原有的PID方法,具有很好的控制效果。Anode baking flue gas temperature is a nonlinear cross-coupling MIMO control object which is main parameter to increase .quality and efficiency of electrolytic aluminum. In order to control the temperature accurately, the control model of the flue gas temperature is identified based on the date gathered from the anode baking furnace scene, which is used for the initial predictive models. A predictive function decoupling control algorithm is proposed, and applied to the muhivariable predictive functional decoupling control for the baking flue gas temperature. The simulation and practical control results show that the proposed control method has better control precision and robust properties than the tradition PID control,and is efficient and effective.

关 键 词：焙烧烟气温度 建模 多变量解耦 预测函数控制 

分 类 号：TP273[自动化与计算机技术—检测技术与自动化装置]