带参数学习的引导信号迭代学习控制方法  

Guiding Signal Iterative Learning Control Method with Parameter Learning

作　　者：黄静[1] 郑华义[2] 李宏 李国岫[1] 邱成[1] HUANG Jing;ZHENG Huayi;LI Hong;LI Guoxiu;QIU Cheng(School of Mechanical,Electronic and Control Engineering,Beijing Jiaotong University,Beijing 100044,China;Beijing Research Institute of Precise Mechatronics and Controls,Beijing 100076,China)

机构地区：[1]北京交通大学机械与电子控制工程学院,北京100044 [2]北京精密机电控制设备研究所,北京100076

出　　处：《兵工学报》2019年第11期2363-2369,共7页Acta Armamentarii

基　　金：中央高校基本科研业务费专项项目(2019RC042)

摘　　要：针对传统迭代学习控制方法应用在空气舵负载模拟器加载系统中的控制发散问题,提出一种引导信号迭代学习控制方法。为提高和增强新方法的智能型和适应性,并具备更快的收敛速度,对新提出方法进行改进:对引导信号进行迭代学习的同时,也对控制参数进行学习,从而使得新控制方法具备双重学习能力。在系统每个迭代周期初始状态不一致的实际情况下,对新控制方法进行收敛特性的数学分析及证明,并最终给出收敛的充分条件。改进后的控制方法应用在空气舵负载模拟器加载系统仿真模型中进行验证,通过与传统迭代学习控制方法以及不带参数学习的控制方法进行对比,表明带参数学习的引导信号迭代学习控制方法具有更快的收敛速度和更好的控制效果。An iterative learning control method for guiding signals is proposed to solve the control divergence in using the traditional iterative learning control method in the loading system of air rudder load simulator.The proposed method is improved for improving its intelligence and adaptability and having a faster convergence speed.The control parameters are learned while the iterative learning of the pilot signal is performed,so the proposed control method has dual learning capabilities.Under the actual situation that the initial state of each iteration cycle of the system is inconsistent,the convergence characteristics of the control method are mathematically analyzed and proven,and the sufficient conditions for convergence are finally given.The improved control method is applied to the simulation model of air rudder load simulator loading system for simulation and verification.Compared with the traditional iterative learning control method and the control method without parameter learning,the proposed iterative learning control method with parameter learning has faster convergence speed and better control effect.

关 键 词：引导信号 迭代学习控制 参数学习 电液伺服系统 

分 类 号：TJ765.42[兵器科学与技术—武器系统与运用工程]