State feedback controller sparsification via a notion of non-fragility
State feedback controller sparsification via a notion of non-fragility
复制标题
通过非脆弱性概念进行状态反馈控制器稀疏化
DOI：
--
复制
发表时间：
2017
期刊：
影响因子：
--
通讯作者：
N. Motee
中科院分区：
文献类型：
--
作者：
MirSaleh Bahavarnia;C. Somarakis;N. Motee
本文引入非脆弱性的概念，用于稳定线性时不变系统的状态反馈控制器。然后，导出了这种非脆弱性的下界和上界。在此基础上，提出了一种稀疏化方法来得到稀疏化状态反馈控制器。通过各种数值实验表明，该方法适用于由数千个状态组成的大系统。此外，在一些案例研究中显示，基于（非脆弱性）的稀疏化过程在稀疏性-性能权衡行为方面优于一种广受尊敬的现有方法。同时，考虑一组稀疏稳定状态反馈控制器，可视化了这种状态反馈控制器的非易损性上界和稀疏性上界之间的权衡。
In this paper, a notion of non-fragility is introduced for a state feedback controller which stabilizes a linear time-invariant (LTI) system. Then, lower and upper bounds on such a non-fragility are derived. Based on such derived bounds on non-fragility, a sparsification procedure is proposed to get sparsified state feedback controllers. Investigating the various numerical experiments, it is observed the proposed method is applicable to large-scale systems consisting of thousands of states. Additionally, it is shown in some case studies, the (non-fragilty)-based sparsification procedure outperforms a well-respected existing method in terms of sparsity-performance tradeoff behavior. Also, considering a set of sparse stabilizing state feedback controllers, a tradeoff between upper bound on non-fragility and sparsity level of such state feedback controllers is visualized.