BP神经网络PID控制器在聚合釜温控中的应用  

Application of BP neural network PID controller in temperature control of polymerizer

作　　者：任有志[1] 杨亚威[1] 曾谦[1] 李竹青[1] 朱家兴[2] 

机构地区：[1]河北科技大学机械工程学院,河北石家庄050000 [2]河北省机电一体化中试基地,河北石家庄050000

出　　处：《现代化工》2016年第2期165-167,169,共4页Modern Chemical Industry

摘　　要：介绍了基于BP神经网络PID控制器的算法,在此基础上提出了多模态神经网络-PID方法。将神经网络PID与普通PID控制方法进行比较,结果表明,该方法具有自适应能力强、过渡时间短、鲁棒性好等特点,弥补了常规PID在聚合釜控制中参数难以整定等缺点。The algorithm of BP neural network PID controller is introduced. The multi-model neural network-PID method is put forward. The comparison of neural network PID with conventional PID control method is performed as well. The results show that this method has the characteristics of strong adaptability, short transition time and good robustness. It has made up for the shortcoming of the difficulty to set in the conventional PID eontrollment of the polymerizer.

关 键 词：BP神经网络 PID控制器 多模态 鲁棒性 

分 类 号：TP13[自动化与计算机技术—控制理论与控制工程]