基于智能模糊神经网络的压力机PID控制器设计与实现  

Design and Implementation of PID Controller Based on Intelligent Fuzzy Neural Network Press

作　　者：杨娟[1] 龚建新[2] YANG Juan;GONG Jianxin(Baise Vocational College,Baise 533000,China;Liuzhou City Vocational College,Liuzhou 545036,China)

机构地区：[1]百色职业学院,广西百色533000 [2]柳州城市职业学院,广西柳州545036

出　　处：《广西民族大学学报（自然科学版）》2020年第2期87-90,95,共5页Journal of Guangxi Minzu University ：Natural Science Edition

基　　金：广西高校教师基础能力提升课题(2019KY1432);广西高校中青年教师能力提升课题(2019KY1278);广西教育厅科研课题(2020KY47008).

摘　　要：针对闭式压力机数学模型构建难度大,控制效果不理想等相关问题,提出一种智能模糊神经网络算法,利用神经网络的记忆特点自动选择各个控制阶段的控制策略,其中模糊算法降低系统的时变性与耦合性,神经网络的自学习功能可以实时对PID参数进行快速准确的调整.仿真结果证明,笔者提出的算法可以明显改善系统的控制性能,与实际环境中的数学模型拟合效果好,大大提高了系统的鲁棒性和灵敏性.The mathematical model of closed press is difficult to construct,and the control effect is not ideal.This paper presents an intelligent fuzzy neural network algorithm using memory characteristics of neural network control strategy to automatically select each control period in which the fuzzy algorithm reduces time-varying systems and coupling neural network self-learning function capable of real-time PID param-eters quickly accurate adjustment.Simulation results prove that proposed algorithm can significantly improve the control system performance a good environment and the actual mathematical model fits well.And the algorithm greatly improves the system's robustness and sensitivity.

关 键 词：闭式压力机 神经网络 PID 时变性 耦合性 仿真实验 

分 类 号：TP183[自动化与计算机技术—控制理论与控制工程]