2014 American Control Conference 2014
DOI: 10.1109/acc.2014.6859227
|View full text |Cite
|
Sign up to set email alerts
|

On the submodularity of sensor scheduling for estimation of linear dynamical systems

Help me understand this report

Search citation statements

Order By: Relevance

Paper Sections

Select...
2
2
1

Citation Types

0
7
0

Year Published

2014
2014
2017
2017

Publication Types

Select...
3
2
1

Relationship

1
5

Authors

Journals

citations
Cited by 8 publications
(7 citation statements)
references
References 14 publications
0
7
0
Order By: Relevance
“…We focus on the case where r = [1 · · · 1] T and β = p for some p ∈ N (i.e., our goal is to choose p sensors out of the total q sensors to optimize the performance of the Kalman filter). In [3], the authors showed that the cost function associated with the singlestep sensor scheduling problem is submodular and thus the greedy algorithm provides a near optimal solution. Thus, in this section, we study a DARE based greedy algorithm for sensor selection, given as Algorithm 1, which iteratively picks sensors that provide the largest incremental decrease in the steady state error covariance.…”
Section: Greedy Algorithmsmentioning
confidence: 99%
See 4 more Smart Citations
“…We focus on the case where r = [1 · · · 1] T and β = p for some p ∈ N (i.e., our goal is to choose p sensors out of the total q sensors to optimize the performance of the Kalman filter). In [3], the authors showed that the cost function associated with the singlestep sensor scheduling problem is submodular and thus the greedy algorithm provides a near optimal solution. Thus, in this section, we study a DARE based greedy algorithm for sensor selection, given as Algorithm 1, which iteratively picks sensors that provide the largest incremental decrease in the steady state error covariance.…”
Section: Greedy Algorithmsmentioning
confidence: 99%
“…3 Note that maximizing F 1 is equivalent to minimizing −F 1 as in the KFSS problem. In [3], the authors showed that the metric F 2 is submodular for the single-step sensor scheduling problem while F 1 and F 3 are neither submodular nor supermodular.…”
Section: A Lack Of Submodularity Of the Cost Functionmentioning
confidence: 99%
See 3 more Smart Citations