密炼机液压上顶栓压力人工神经网络控制系统算法研究  

The emulation research of the Neural Networks control system of the RAM's pressure based on Matlab/simulink

作　　者：陈珍颖 罗春雷[1] 刘健[1] 

机构地区：[1]中南大学机电学院,湖南长沙410083

出　　处：《橡塑技术与装备》2014年第9期1-6,共6页China Rubber/Plastics Technology and Equipment

基　　金：湖南省科技计划项目(07GK3041)

摘　　要：阐述了BP神经网络控制算法原理;论证了人工神经网络相对于经典PID控制在上顶栓压力闭环控制上优越性;建立了系统的机电液耦合模型,通过MATLAB仿真实验,研究BP算法在该系统中的动态特性,为神经网络在该控制系统中的应用提供实验依据;最后仿真结果表明,引入BP算法可以使系统超调量明显下降,响应时间显著得到提高,静差得到很好的消除。Expatiate the principle of BP Neural Networks arithmetic; Argument Neural Networks＇s Superiority relative to PID control on the RAM＇s pressure; Set up the machine-electricity-fluid coupling model; Research the BP arithmetic＇s dynamic characteristic by Matlab to offer prove on experiment to Neural Networks on the control system; The result of emulation indicates that superscale drops, response time increases evidently and static error is eliminate.

关 键 词：BP神经网络 MATLAB 密炼机 上顶栓 压力 动态特性 

分 类 号：TQ330.493[化学工程—橡胶工业]