An Asynchronous Approximate Distributed Alternating Direction Method of Multipliers in Digraphs

被引:6
作者
Jiang, Wei [1 ]
Grammenos, Andreas [2 ,3 ]
Kalyvianaki, Evangelia [2 ]
Charalambous, Themistoklis [1 ]
机构
[1] Aalto Univ, Sch Elect Engn, Dept Elect Engn & Automat, Espoo, Finland
[2] Univ Cambridge, Dept Comp Sci & Technol, Cambridge, England
[3] Alan Turing Inst, London, England
来源
2021 60TH IEEE CONFERENCE ON DECISION AND CONTROL (CDC) | 2021年
基金
芬兰科学院; 英国工程与自然科学研究理事会;
关键词
Distributed optimization; asynchronous ADMM; directed graphs; ratio consensus; finite-time consensus;
D O I
10.1109/CDC45484.2021.9683229
中图分类号
TP [自动化技术、计算机技术];
学科分类号
0812 ;
摘要
In this work, we consider the asynchronous distributed optimization problem in which each node has its own convex cost function and can communicate directly only with its neighbors, as determined by a directed communication topology (directed graph or digraph). First, we reformulate the optimization problem so that Alternating Direction Method of Multipliers (ADMM) can be utilized. Then, we propose an algorithm, herein called Asynchronous Approximate Distributed Alternating Direction Method of Multipliers (AsyAD-ADMM), using finite-time asynchronous approximate ratio consensus, to solve the multi-node convex optimization problem, in which every node performs iterative computations and exchanges information with its neighbors asynchronously. More specifically, at every iteration of AsyAD-ADMM, each node solves a local convex optimization problem for the one of the primal variables and utilizes a finite-time asynchronous approximate consensus protocol to obtain the value of the other variable which is close to the optimal value, since the cost function for the second primal variable is not decomposable. If the individual cost functions are convex, but not-necessarily differentiable, the proposed algorithm converges at a rate of O(1/k), where k is the iteration counter. The efficacy of AsyAD-ADMM is exemplified via a proof-of-concept distributed least square optimization problem with different performance-influencing factors investigated.
引用
收藏
页码:3406 / 3413
页数:8
相关论文
共 18 条
  • [1] Asynchronous Distributed Optimization Over Lossy Networks via Relaxed ADMM: Stability and Linear Convergence
    Bastianello, Nicola
    Carli, Ruggero
    Schenato, Luca
    Todescato, Marco
    [J]. IEEE TRANSACTIONS ON AUTOMATIC CONTROL, 2021, 66 (06) : 2620 - 2635
  • [2] Distributed optimization and statistical learning via the alternating direction method of multipliers
    Boyd S.
    Parikh N.
    Chu E.
    Peleato B.
    Eckstein J.
    [J]. Foundations and Trends in Machine Learning, 2010, 3 (01): : 1 - 122
  • [3] Finite-Time Approximate Consensus and its Application to Distributed Frequency Regulation in Islanded AC Microgrids
    Cady, Stanton T.
    Dominguez-Garcia, Alejandro D.
    Hadjicostis, Christoforos N.
    [J]. 2015 48TH HAWAII INTERNATIONAL CONFERENCE ON SYSTEM SCIENCES (HICSS), 2015, : 2664 - 2670
  • [4] Asynchronous Distributed ADMM for Large-Scale Optimization-Part I: Algorithm and Convergence Analysis
    Chang, Tsung-Hui
    Hong, Mingyi
    Liao, Wei-Cheng
    Wang, Xiangfeng
    [J]. IEEE TRANSACTIONS ON SIGNAL PROCESSING, 2016, 64 (12) : 3118 - 3130
  • [5] Distributed Finite-Time Computation of Digraph Parameters: Left-Eigenvector, Out-Degree and Spectrum
    Charalambous, Themistoklis
    Rabbat, Michael G.
    Johansson, Mikael
    Hadjicostis, Christoforos N.
    [J]. IEEE TRANSACTIONS ON CONTROL OF NETWORK SYSTEMS, 2016, 3 (02): : 137 - 148
  • [6] Dominguez-Garcia AD, 2010, INT CONF SMART GRID, P537, DOI 10.1109/SMARTGRID.2010.5621991
  • [7] Giannini S, 2013, IEEE DECIS CONTR P, P2605, DOI 10.1109/CDC.2013.6760275
  • [8] Grammenos A., 2021, ARXIV210106139
  • [9] Average Consensus in the Presence of Delays in Directed Graph Topologies
    Hadjicostis, Christoforos N.
    Charalambous, Themistoklis
    [J]. IEEE TRANSACTIONS ON AUTOMATIC CONTROL, 2014, 59 (03) : 763 - 768
  • [10] Jiang W., 2021, P 19 EUR CONTR C JUN, P2198