摘要
提出了一种基于神经网络最小参数学习法的RBF网络自适应鲁棒滑模控制方法,在设定预期轨迹的前提下,利用RBF神经网络对未知参数进行自适应逼近,通过参数估计代替神经网络权值的调整,加快了自适应律的求解。在空载、半载、满载3种情况下的仿真结果表明,该控制算法对系统参数的大范围变化和外界的不稳定扰动可以进行自适应调整,具有较好的学习性能和控制精度。
This paper presents an adaptive robust sliding mode control method based on neural network minimum parameter learning method for RBF neural network.After setting the expected trajectory,using RBF neural networks to approximate unknown parameters,we replaced the adjustment of neural network weights by parameter estimation,accelerated the solution of adaptive laws.The simulation resultsshow that,the control algorithm which proposed in this paper has excellent learning performance and control accuracy under three conditions of no load,half load and full load,and it can make adjustments for changes in system parameters and unstable disturbance.
作者
唐逸雄
陈龙淼
高波
TANG Yixiong;CHEN Longmiao;GAO Bo(School of Mechanical Engineering,Nanjing University of Science and Technology,Nanjing 210094,China;Army Academy of Artillery Air Defense Research Institute,Beijing 100012,China)
出处
《兵器装备工程学报》
CAS
北大核心
2018年第11期135-139,共5页
Journal of Ordnance Equipment Engineering
关键词
兵器科学与技术
自动装填系统
链式回转弹仓
RBF神经网络
滑模控制
weapons science and technology
automatic loading system
rotational chain shell magazine
RBF neural networks
sliding mode control
