摘要
运用最优方法探讨了一类模型不确定系统的鲁棒控制问题。定义了一个实际敏感度和标称敏感度之间的加权敏感度误差,并通过调整标称控制器最小化加权敏感度误差在整个频段上的方差,从而为一类不确定系统提供了一种最优的鲁棒控制器设计方法,可使系统性能对模型误差具有良好的鲁棒性。针对控制工程领域典型的一阶时滞系统进行仿真研究,其结果说明了该方法的有效性。
Optimization method is applied to the optimal robust control for a class of uncertain svstems. A weighed sensitivity error between actual sensitivity and nominal sensitivity is defined. By modifying the nominal controller to minimize the variance of the weighed sensitivity error over all frequency, an optimal robust controller design method is presented for a class of tmcertain system, which can achieve good robustness against model errors. The simulation research for typical one-order and time-delay system in the control engineering application show the validity of the obtained method.
出处
《控制工程》
CSCD
2008年第1期53-56,共4页
Control Engineering of China
基金
国家863项目基金资助项目(2003AA517020)
关键词
最优鲁棒控制
模型不确定性
谱分解
optimal robust control
plant uncertainty
spectral factorization
