神经滑模控制在转台伺服系统应用  

Application of neural network sliding mode control for turntable servo system

作　　者：张飞云[1] 王武[1] 姚宁[1] 

机构地区：[1]许昌学院电气信息工程学院,许昌461000

出　　处：《机械设计与制造》2011年第9期65-67,共3页Machinery Design & Manufacture

基　　金：河南省教育厅自然科学研究资助项目(2011B470005);河南省科技厅科技攻关计划资助项目(112102210339)

摘　　要：以直流电机为执行机构,分析了飞行仿真转台伺服系统的数学模型。滑模控制具有对系统扰动和参数摄动的自适应性,可实现伺服系统的快速响应,同时有效克服低速状态下摩擦力矩的影响。系统抖振问题是滑模控制的突出问题,利用径向基神经网络的非线性逼近能力,给出以切换函数为网络输入,以滑模控制器为网络输出,构建了神经滑模控制器,软件仿真结果表明所设计的滑模控制器能达到较好的控制品质,有效的克服系统抖振和外部扰动,实现系统低速摩擦补偿。The mathematical model of the flight simulator servo system was analyzed based on DC motor.Sliding mode control is an adaptive control strategy which is suitable for servo system for its quick re-sponse and friction torque elimination at low speed.As the chattering is the key problem of SMC and the neural network sliding mode controller is constructed with switch function as the input of neural network and sliding mode controller as the output.The simulation shows the sliding mode controller designed achieves better control performance,which reduces chattering and disturbance effectively,and low speed friction compensation is achieved.

关 键 词：神经滑模控制 伺服转台 径向基神经网络型 仿真 

分 类 号：TH16[机械工程—机械制造及自动化] TP273[自动化与计算机技术—检测技术与自动化装置]