Efficient control recovery for resilient control systems

Shuo Zhang, Stephen Wolthusen

Research output: Contribution to conferencePaperpeer-review

125 Downloads (Pure)

Abstract

Resilient control systems should efficiently restore control into physical systems not only after the sabotage of themselves, but also after breaking physical systems. To enhance resilience of control systems, and given an originally minimal-input controlled linear-time invariant(LTI) physical system, we address the problem of efficient control recovery into it after removing a known system vertex by finding the minimum number of inputs. According to the minimum input theorem, with a digraph embedded into LTI model and involving a precomputed maximum matching, this problem is modeled into recovering controllability of it after removing a known network vertex. Then, we recover controllability of the residual network by efficiently finding a maximum matching rather than recomputation. As a result, except for precomputing a maximum matching and predetermining the removed vertex, the worst-case execution time of control recovery into the residual LTI physical system is linear.
Original languageEnglish
Pages1-6
Number of pages6
DOIs
Publication statusPublished - 21 May 2018
Event15th IEEE International Conference on Networking, Sensing and Control - Zhuhai , Zhuhai, China
Duration: 27 Mar 201829 Mar 2018
https://icnsc2018.jnu.edu.cn/

Conference

Conference15th IEEE International Conference on Networking, Sensing and Control
Country/TerritoryChina
CityZhuhai
Period27/03/1829/03/18
Internet address

Keywords

  • Resilient Control
  • Control recovery
  • Maximum Cardinality Matching

Cite this