弹仓的RBF神经网络自适应滑模控制  

RBF Neural Network Adaptive Sliding Mode Control of Magazine

作　　者：王思杰 李志刚[1] 孙玲庆 WANG Sijie;LI Zhigang;SUN Lingqing(Mechanical Engineering College, Nanjing University of Science and Technology, Nanjing 210094, China)

机构地区：[1]南京理工大学机械工程学院,南京210094

出　　处：《兵器装备工程学报》2021年第4期208-212,共5页Journal of Ordnance Equipment Engineering

摘　　要：提出了一种智能算法与滑模控制相结合的控制策略,通过自适应算法对弹仓系统未知参数进行估计,通过RBF神经网络逼近滑模控制中的切换项,形成无模型控制器。在弹仓空载、半载和满载3种不同情况下进行仿真验证,结果表明,该控制算法对弹仓系统中存在的非线性摩擦力矩、啮合冲击力矩、链传动多边形效应以及其他未知的外部扰动力矩等不敏感,具有较好的鲁棒性,控制精度较高。A control strategy combining intelligent algorithm and sliding mode control was proposed.The unknown parameters of the magazine system were estimated by adaptive algorithm,and then the switching term in sliding mode control was approached by RBF neural network to form a model free controller.Simulation verification was carried out under three different conditions:no load,half load and full load.The simulation results show that the control algorithm is insensitive to the nonlinear friction moment,meshing impact moment,chain drive polygon effect and other unknown external disturbance moment in the magazine system,and has good robustness and high control accuracy.

关 键 词：自动装填系统 滑模控制 自适应算法 RBF神经网络 自动化弹仓 

分 类 号：TP273[自动化与计算机技术—检测技术与自动化装置]