基于神经网络PID的连挤连轧速度控制系统  

Speed Control System Based on the Continuous Extrusion and Rolling of Neural Network PID

作　　者：刘富东[1] 祁伟[1] 运新兵[1] 

机构地区：[1]大连交通大学连续挤压教育部工程研究中心,辽宁大连116028

出　　处：《大连交通大学学报》2014年第5期92-94,120,共4页Journal of Dalian Jiaotong University

基　　金：国家自然科学基金资助项目(51175055);国家科技支撑计划资助项目(2011BAEr3B01);辽宁省自然科学基金资助项目(201102020)

摘　　要：根据连挤连轧速度控制的特点,将改进BP神经网络与PID控制算法相结合,采用西门子S7-200型PLC实现连挤连轧机的变频调速控制.通过与传统PID和BP神经网络PID算法比较,改进BP神经网络PID控制器具有较短的动态响应时间和良好的跟随性,表明该方法在连挤连轧速度控制中有良好的应用价值.Based on the characteristics of continuous extrusion and rolling speed control, the improved BP neural network is combined with PID control algorithm, and Siemens S7-200 PLC is adopted to realize the tan- dem mill of motor frequency and speed controls. By comparison with traditional PID and BP neural network PID algorithm, the improved BP neural network PID controller has shorter dynamic response time and good fol- lowing characteristics, which indicates that the method has a good application value in the crowded and rolling speed control.

关 键 词：连挤连轧 改进BP神经网络 速度控制 

分 类 号：TG334.9[金属学及工艺—金属压力加工]