A Decentralized Second-Order Method for Dynamic Optimization

被引:0
作者
Mokhtari, Aryan [1 ]
Shi, Wei [2 ]
Ling, Qing [3 ]
Ribeiro, Alejandro [1 ]
机构
[1] Univ Penn, Dept Elect & Syst Engn, Philadelphia, PA 19104 USA
[2] Univ Illinois, Coordinated Sci Lab, 1308 W Main St, Urbana, IL 61801 USA
[3] Univ Sci & Technol China, Dept Automat, 96 Jinzhao Rd, Hefei 230026, Anhui, Peoples R China
来源
2016 IEEE 55TH CONFERENCE ON DECISION AND CONTROL (CDC) | 2016年
关键词
multi-agent network; decentralized optimization; dynamic optimization; second-order methods; ALTERNATING DIRECTION METHOD; CONSENSUS;
D O I
暂无
中图分类号
TP [自动化技术、计算机技术];
学科分类号
0812 ;
摘要
This paper considers decentralized dynamic optimization problems where nodes of a network try to minimize a sequence of time-varying objective functions in a real-time scheme. At each time slot, nodes have access to different summands of an instantaneous global objective function and they are allowed to exchange information only with their neighbors. This paper develops the application of the Exact Second-Order Method (ESOM) to solve the dynamic optimization problem in a decentralized manner. The proposed dynamic ESOM algorithm operates by primal descending and dual ascending on a quadratic approximation of an augmented Lagrangian of the instantaneous consensus optimization problem. The convergence analysis of dynamic ESOM indicates that a Lyapunov function of the sequence of primal and dual errors converges linearly to an error bound when the local functions are strongly convex and have Lipschitz continuous gradients. Numerical results demonstrate the claim that the sequence of iterates generated by the proposed method is able to track the sequence of optimal arguments.
引用
收藏
页码:6036 / 6043
页数:8
相关论文
共 29 条
[11]   Adaptive Information Collection by Robotic Sensor Networks for Spatial Estimation [J].
Graham, Rishi ;
Cortes, Jorge .
IEEE TRANSACTIONS ON AUTOMATIC CONTROL, 2012, 57 (06) :1404-1419
[12]  
Hestenes M. R., 1969, Journal of Optimization Theory and Applications, V4, P303, DOI 10.1007/BF00927673
[13]   Linear Convergence Rate of a Class of Distributed Augmented Lagrangian Algorithms [J].
Jakovetic, Dusan ;
Moura, Jose M. F. ;
Xavier, Joao .
IEEE TRANSACTIONS ON AUTOMATIC CONTROL, 2015, 60 (04) :922-936
[14]   Fast Distributed Gradient Methods [J].
Jakovetic, Dusan ;
Xavier, Joao ;
Moura, Jose M. F. .
IEEE TRANSACTIONS ON AUTOMATIC CONTROL, 2014, 59 (05) :1131-1146
[15]   Cooperative Convex Optimization in Networked Systems: Augmented Lagrangian Algorithms With Directed Gossip Communication [J].
Jakovetic, Dusan ;
Xavier, Joao ;
Moura, Jose M. F. .
IEEE TRANSACTIONS ON SIGNAL PROCESSING, 2011, 59 (08) :3889-3902
[16]   D-MAP: Distributed Maximum a Posteriori Probability Estimation of Dynamic Systems [J].
Jakubiec, Felicia Y. ;
Ribeiro, Alejandro .
IEEE TRANSACTIONS ON SIGNAL PROCESSING, 2013, 61 (02) :450-466
[17]   Gossip and Distributed Kalman Filtering: Weak Consensus Under Weak Detectability [J].
Kar, Soummya ;
Moura, Jose M. F. .
IEEE TRANSACTIONS ON SIGNAL PROCESSING, 2011, 59 (04) :1766-1784
[18]   Decentralized Dynamic Optimization Through the Alternating Direction Method of Multipliers [J].
Ling, Qing ;
Ribeiro, Alejandro .
IEEE TRANSACTIONS ON SIGNAL PROCESSING, 2014, 62 (05) :1185-1197
[19]  
Mokhtari A., 2016, ARXIV160200596
[20]   DQM: Decentralized Quadratically Approximated Alternating Direction Method of Multipliers [J].
Mokhtari, Aryan ;
Shi, Wei ;
Ling, Qing ;
Ribeiro, Alejandro .
IEEE TRANSACTIONS ON SIGNAL PROCESSING, 2016, 64 (19) :5158-5173