| 引用本文: | 刘聪,李颖晖,朱喜华,刘培培.基于自适应滑模观测器的不匹配非线性系统执行器故障重构[J].控制理论与应用,2014,31(4):431~437.[点击复制] |
| LIU Cong,LI Ying-hui,ZHU Xi-hua,LIU Pei-pei.Adaptive sliding-mode observer for actuator fault reconstruction in nonlinear system with mismatched uncertainties[J].Control Theory and Technology,2014,31(4):431~437.[点击复制] |
|
| 基于自适应滑模观测器的不匹配非线性系统执行器故障重构 |
| Adaptive sliding-mode observer for actuator fault reconstruction in nonlinear system with mismatched uncertainties |
| 摘要点击 2787 全文点击 2175 投稿时间：2013-09-10 修订日期：2013-12-07 |
| 查看全文 查看/发表评论 下载PDF阅读器 |
| DOI编号 10.7641/CTA.2014.30952 |
| 2014,31(4):431-437 |
| 中文关键词 自适应滑模观测器 线性矩阵不等式 故障重构 高增益观测器 |
| 英文关键词 adaptive sliding-mode observer linear matrix inequality fault reconstruction high gain observer |
| 基金项目 国家自然科学基金资助项目(61074007); 总装备部预研基金资助项目. |
|
| 中文摘要 |
| 针对一类含有未知干扰的不匹配非线性Lipschitz系统, 提出了基于自适应滑模观测器的执行器故障重构方法. 首先引入辅助输出矩阵, 使得辅助输出系统的观测器匹配条件得以满足, 同时设计了高增益观测器实现对未知辅助输出的精确估计; 然后针对辅助输出系统建立故障重构滑模观测器, 设计了自适应律在线修正滑模控制器增益, 考虑故障上界未知的前提下, 提出了观测器状态估计误差稳定的存在定理, 运用Schur补引理将观测器反馈增益矩阵设计方法转化为求解线性矩阵不等式约束优化问题, 同时引入线性变换矩阵, 在故障上界未知的前提下设计了滑模控制增益, 使得输出估计误差收敛稳定, 确保了滑模运动在有限时间内发生, 在此基础上利用等效控制输出误差注入原理实现了执行器故
障重构; 最后通过仿真算例验证了本文方法的有效性. |
| 英文摘要 |
| An estimator of actuator fault is proposed, based on adaptive sliding-mode observer for the unmatched Lipschitz nonlinear system with unknown disturbances. Firstly, an auxiliary matrix of output is introduced, which makes the matching condition satisfied; and a high gain observer is designed, which is used for the accurate estimation of the unknown auxiliary output. Secondly, a sliding-mode observer for fault reconstruction is designed for the auxiliary output system and an adaptive law is designed to modify the gain of the controller on-line by considering the condition of unknown upper bound fault; a theorem for the stabilization of the estimated state error is put forward; and under the application of complementary lemma of Schur, the designed method of observer feedback matrix is transformed to an optimization issue and is solved by the linear matrix inequality; meanwhile, a linear transformed matrix is introduced; the sliding-mode gain is designed under the unknown upper fault bound, which makes the estimate error of output stable and the sliding-mode motion take place in finite time; then, under the application of equivalent output injection concept, the actuator fault can be reconstructed. Finally, an example validates the effectiveness of the proposed method. |
|
|
|
|
|