| 引用本文: | 孙宜标, 郭庆鼎.基于RBF神经网络补偿的直线伺服系统滑模鲁棒跟踪控制[J].控制理论与应用,2004,21(2):252~256.[点击复制] |
| SUN Yi-biao, GUO Qing-ding.Sliding mode robust tracking control for linear servo system based on RBF neural networks compensation[J].Control Theory and Technology,2004,21(2):252~256.[点击复制] |
|
| 基于RBF神经网络补偿的直线伺服系统滑模鲁棒跟踪控制 |
| Sliding mode robust tracking control for linear servo system based on RBF neural networks compensation |
| 摘要点击 1125 全文点击 1646 投稿时间：2002-03-20 修订日期：2003-05-15 |
| 查看全文 查看/发表评论 下载PDF阅读器 |
| DOI编号 |
| 2004,21(2):252-256 |
| 中文关键词 永磁直线同步电机 直接驱动 端部效应 滑模变结构控制 抖振 径向基函数神经网络 |
| 英文关键词 permanent-magnet linear synchronization motor direct-drive end effects sliding mode variable structure control chattering radial basis function neuron network |
| 基金项目 国家自然科学基金项目(50075057); 辽宁省教育厅基金项目(20141040). |
|
| 中文摘要 |
| 永磁直线伺服系统具有高速、高响应和直接驱动等优点,但负载扰动、端部效应、非线性摩擦及系统参数变化会降低系统的伺服性能.为了在保证系统的跟踪性能的基础上.消除上述不确定性因素的影响,本文提出一种将变结构控制(VSC)和径向基函数神经网络(RBFNN)相结合的鲁棒跟踪控制策略.变结构控制具有快速响应,对不确定因素的不变性的优点.但是其“抖振”现象将影响直线伺服系统的平稳性和定位精度.采用径向基函数神经网络来模拟端部效应、参数变化、摩擦和外部负载等不确定因素,引入带死区的目标函数以缩短学习过程.通过RBFNN的补偿控制来减弱“抖振”输入的程度,进一步提高系统的稳态精度.仿真结果表明,该方案对直线伺服系统不确定性有很强的鲁棒性,同时,系统具有较好的跟踪性能,大大提高了直接驱动直线伺服系统的鲁棒跟踪精度. |
| 英文摘要 |
| Permanent-magnet linear servo system has the merits of high speed, high response, and direct drive etc., but the load disturbance, end effects, nonlinear friction, and the change of system parameters reduce the servo performance of the system. To eliminate the influence of the uncertainties mentioned above for ensuring tracking capability, in this paper a robust tracking control strategy is proposed, combining the variable structure control (VSC) with the radial basis function neuron network (RBFNN). The VSC has the merits of high response and the invariability to uncertainties, but its "chattering" phenomenon negatively affects the placidity and positioning precision of the linear servo system. An RBFNN is applied to model the uncertainties caused by end effects, parameter variations, friction, and external load etc., and an objective function with dead zone is introduced to shorten the learning process. The compensation control based on RBFNN attenuates the chattering level of the control input and improves the static precision of the system. The simulation results show that this control scheme not only has a strong robustness to uncertainties of the linear system, but also has a good tracking performance. In fact, the control greatly improves the robust tracking precision of the direct drive linear servo system. |
|
|
|
|
|