Accelerated gradient methods and dual decomposition in distributed model predictive control
(2013) In Automatica 49(3). p.829-833- Abstract
- We propose a distributed optimization algorithm for mixed
L_1/L_2-norm optimization based on accelerated gradient methods using dual decomposition. The algorithm achieves convergence rate O(1/k^2), where k is the iteration number, which significantly improves the convergence rates of existing duality-based distributed optimization algorithms that achieve O(1/k). The performance of the developed algorithm is evaluated on randomly generated optimization problems arising in distributed model predictive control (DMPC). The evaluation shows that, when the problem data is sparse and large-scale, our algorithm can outperform current state-of-the-art optimization software CPLEX and MOSEK.
Please use this url to cite or link to this publication:
https://lup.lub.lu.se/record/2278086
- author
- Giselsson, Pontus LU ; Doan, Dang ; Keviczky, Tamas ; De Schutter, Bart and Rantzer, Anders LU
- organization
- publishing date
- 2013
- type
- Contribution to journal
- publication status
- published
- subject
- in
- Automatica
- volume
- 49
- issue
- 3
- pages
- 829 - 833
- publisher
- Pergamon Press Ltd.
- external identifiers
-
- wos:000316590300017
- scopus:84875216516
- ISSN
- 0005-1098
- DOI
- 10.1016/j.automatica.2013.01.009
- project
- LCCC
- language
- English
- LU publication?
- yes
- additional info
- key= gis2012aut
- id
- ee512043-aeca-4559-bf78-5cf1cbc81547 (old id 2278086)
- date added to LUP
- 2016-04-01 14:31:58
- date last changed
- 2024-05-09 01:18:16
@article{ee512043-aeca-4559-bf78-5cf1cbc81547, abstract = {{We propose a distributed optimization algorithm for mixed<br/><br> L_1/L_2-norm optimization based on accelerated gradient methods using dual decomposition. The algorithm achieves convergence rate O(1/k^2), where k is the iteration number, which significantly improves the convergence rates of existing duality-based distributed optimization algorithms that achieve O(1/k). The performance of the developed algorithm is evaluated on randomly generated optimization problems arising in distributed model predictive control (DMPC). The evaluation shows that, when the problem data is sparse and large-scale, our algorithm can outperform current state-of-the-art optimization software CPLEX and MOSEK.}}, author = {{Giselsson, Pontus and Doan, Dang and Keviczky, Tamas and De Schutter, Bart and Rantzer, Anders}}, issn = {{0005-1098}}, language = {{eng}}, number = {{3}}, pages = {{829--833}}, publisher = {{Pergamon Press Ltd.}}, series = {{Automatica}}, title = {{Accelerated gradient methods and dual decomposition in distributed model predictive control}}, url = {{https://lup.lub.lu.se/search/files/4027400/3131655.pdf}}, doi = {{10.1016/j.automatica.2013.01.009}}, volume = {{49}}, year = {{2013}}, }