Distributed event-triggered algorithm for optimal resource allocation of multi-agent systems

Weiyong Yu; Zhenhua Deng; Hongbing Zhou; Xianlin Zeng

Kybernetika (2017)

  • Volume: 53, Issue: 5, page 747-764
  • ISSN: 0023-5954

Abstract

top
This paper is concerned with solving the distributed resource allocation optimization problem by multi-agent systems over undirected graphs. The optimization objective function is a sum of local cost functions associated to individual agents, and the optimization variable satisfies a global network resource constraint. The local cost function and the network resource are the private data for each agent, which are not shared with others. A novel gradient-based continuous-time algorithm is proposed to solve the distributed optimization problem. We take an event-triggered communication strategy and an event-triggered gradient measurement strategy into account in the algorithm. With strongly convex cost functions and locally Lipschitz gradients, we show that the agents can find the optimal solution by the proposed algorithm with exponential convergence rate, based on the construction of a suitable Lyapunov function. Finally, a numerical example is provided to demonstrate the effectiveness of the proposed scheme.

How to cite

top

Yu, Weiyong, et al. "Distributed event-triggered algorithm for optimal resource allocation of multi-agent systems." Kybernetika 53.5 (2017): 747-764. <http://eudml.org/doc/294290>.

@article{Yu2017,
abstract = {This paper is concerned with solving the distributed resource allocation optimization problem by multi-agent systems over undirected graphs. The optimization objective function is a sum of local cost functions associated to individual agents, and the optimization variable satisfies a global network resource constraint. The local cost function and the network resource are the private data for each agent, which are not shared with others. A novel gradient-based continuous-time algorithm is proposed to solve the distributed optimization problem. We take an event-triggered communication strategy and an event-triggered gradient measurement strategy into account in the algorithm. With strongly convex cost functions and locally Lipschitz gradients, we show that the agents can find the optimal solution by the proposed algorithm with exponential convergence rate, based on the construction of a suitable Lyapunov function. Finally, a numerical example is provided to demonstrate the effectiveness of the proposed scheme.},
author = {Yu, Weiyong, Deng, Zhenhua, Zhou, Hongbing, Zeng, Xianlin},
journal = {Kybernetika},
keywords = {distributed optimization; event-triggered strategy; multi-agent systems; resource allocation},
language = {eng},
number = {5},
pages = {747-764},
publisher = {Institute of Information Theory and Automation AS CR},
title = {Distributed event-triggered algorithm for optimal resource allocation of multi-agent systems},
url = {http://eudml.org/doc/294290},
volume = {53},
year = {2017},
}

TY - JOUR
AU - Yu, Weiyong
AU - Deng, Zhenhua
AU - Zhou, Hongbing
AU - Zeng, Xianlin
TI - Distributed event-triggered algorithm for optimal resource allocation of multi-agent systems
JO - Kybernetika
PY - 2017
PB - Institute of Information Theory and Automation AS CR
VL - 53
IS - 5
SP - 747
EP - 764
AB - This paper is concerned with solving the distributed resource allocation optimization problem by multi-agent systems over undirected graphs. The optimization objective function is a sum of local cost functions associated to individual agents, and the optimization variable satisfies a global network resource constraint. The local cost function and the network resource are the private data for each agent, which are not shared with others. A novel gradient-based continuous-time algorithm is proposed to solve the distributed optimization problem. We take an event-triggered communication strategy and an event-triggered gradient measurement strategy into account in the algorithm. With strongly convex cost functions and locally Lipschitz gradients, we show that the agents can find the optimal solution by the proposed algorithm with exponential convergence rate, based on the construction of a suitable Lyapunov function. Finally, a numerical example is provided to demonstrate the effectiveness of the proposed scheme.
LA - eng
KW - distributed optimization; event-triggered strategy; multi-agent systems; resource allocation
UR - http://eudml.org/doc/294290
ER -

References

top
  1. Beck, A., Nedić, A., Ozdaglar, A., Teboulle, M., 10.1109/tcns.2014.2309751, IEEE Trans. Control Network Systems 1 (2014), 64-74. MR3245002DOI10.1109/tcns.2014.2309751
  2. Chen, W., Ren, W., 10.1016/j.automatica.2015.11.015, Automatica 65 (2016), 90-97. Zbl1328.93167MR3447697DOI10.1016/j.automatica.2015.11.015
  3. Deng, Z., Wang, X., Hong, Y., 10.1049/iet-cta.2016.0795, IET Control Theory Appl. 11 (2017), 282-290. MR3675134DOI10.1049/iet-cta.2016.0795
  4. Ghadimi, E., Shames, I., Johansson, M., 10.1109/tsp.2013.2278149, IEEE Trans. Signal Processing 61 (2013), 5417-5429. DOI10.1109/tsp.2013.2278149
  5. Godsil, C., Royle, G., 10.1007/978-1-4613-0163-9, Springer-Verlag, New York 2001. Zbl0968.05002MR1829620DOI10.1007/978-1-4613-0163-9
  6. Gao, Y., Wang, L., 10.1109/tac.2011.2112472, IEEE Trans. Automat. Control 56 (2011), 1226-1231. MR2815935DOI10.1109/tac.2011.2112472
  7. Ho, Y. C., Servi, L., Suri, R., A class of center-free resource allocation algorithms., Large Scale Systems 1 (1980), 51-62. MR0617154
  8. Khalil, H. K., Nonlinear Systems. Third edition., Prentice Hall, New Jersey 2002. 
  9. Kia, S. S., Cortés, J., Martínez, S., 10.1016/j.automatica.2015.03.001, Automatica 55 (2015), 254-264. MR3336675DOI10.1016/j.automatica.2015.03.001
  10. Lehmann, D., Lunze, J., 10.1080/00207179.2012.659760, Int. J. Control 85 (2012), 563-577. MR2913571DOI10.1080/00207179.2012.659760
  11. Lakshmanan, H., Farias, D. P., 10.1137/060662228, SIAM J. Optim. 19 (2008), 911-940. MR2448920DOI10.1137/060662228
  12. Hu, J., Chen, G., Li, H., Distributed event-triggered tracking control of leader-follower multi-agent systems with communication delays., Kybernetika 47 (2011), 630-643. Zbl1227.93008MR2884865
  13. Nedić, A., Ozdaglar, A., Parrilo, P. A., 10.1109/tac.2010.2041686, IEEE Trans. Automat. Control 55 (2010), 922-938. MR2654432DOI10.1109/tac.2010.2041686
  14. Lou, Y., Shi, G., Johansson, K., Hong, Y., 10.1109/tac.2014.2309261, IEEE Trans. Automat. Control 59 (2014), 1722-1736. MR3232068DOI10.1109/tac.2014.2309261
  15. Rockafellar, R. T., Convex Analysis., Princeton University Press, Princeton 1972. Zbl1011.49013MR0274683
  16. Ruszczynski, A. P., Nonlinear Optimization., Princeton University Press, Princeton 2006. MR2199043
  17. Tabuada, P., 10.1109/tac.2007.904277, IEEE Trans. Automat. Control 52 (2007), 1680-1685. MR2352444DOI10.1109/tac.2007.904277
  18. Wang, X., Yi, P., Hong, Y., 10.1007/s11768-014-0036-y, Control Theory Technol. 12 (2014), 132-138. MR3199533DOI10.1007/s11768-014-0036-y
  19. Wang, X., Deng, Z., Ma, S., Du, X., Event-triggered design for multi-agent optimal consensus of Euler-Lagrangian systems., Kybernetika 1 (2017), 179-194. MR3638563
  20. Xiao, L., Boyd, S., 10.14736/kyb-2017-1-0179, J. Optim. Theory and Appl. 129 (2006), 469-488. MR2281152DOI10.14736/kyb-2017-1-0179
  21. Zhang, Y., Hong, Y., 10.1109/wcica.2012.6358108, In: Proc. 10th IEEE World Congress on Intelligent Control and Automation 2012, pp. 1453-1458. DOI10.1109/wcica.2012.6358108
  22. Zhang, Y., Lou, Y., Hong, Y., Xie, L., 10.1109/twc.2015.2402672, IEEE Trans. Wireless Communications 14 (2015), 3131-3142. MR2960500DOI10.1109/twc.2015.2402672
  23. Yi, P., Hong, Y., Liu, F., 10.1016/j.automatica.2016.08.007, Automatica 74 (2016), 259-269. MR3569392DOI10.1016/j.automatica.2016.08.007

NotesEmbed ?

top

You must be logged in to post comments.

To embed these notes on your page include the following JavaScript code on your page where you want the notes to appear.

Only the controls for the widget will be shown in your chosen language. Notes will be shown in their authored language.

Tells the widget how many notes to show per page. You can cycle through additional notes using the next and previous controls.

    
                

                
Note: Best practice suggests putting the JavaScript code just before the closing </body> tag.