.png){kind=icon}
.png){kind=icon}
Return
Event-Triggered Zero-Gradient-Sum Distributed Constrained Optimization Over Jointly Connected Balanced Digraphs
.png){kind=icon}
PRE
AI
DOI：10.1109/TNSE.2025.3559905.png)
.png){kind=icon}
Original request for help.png){kind=icon}
Share.png){kind=icon}
Save
Abstract
En 中文
In the realm of distributed optimization (DO), it is expected to design a distributed algorithm that has a lower communication burden while handling general constraints over switching graphs. One promising approach is the zero-gradient-sum (ZGS) algorithm. However, existing ZGS-based discrete-time algorithms are limited to unconstrained DO on fixed network structures. This paper addresses this gap by first providing an event-triggered ZGS (ET-ZGS) algorithm for solving equality-constrained DO over uniformly jointly strongly connected (UJSC) and balanced digraphs. Sufficient conditions on the fixed step size are derived to guarantee the convergence for switching graphs. Specifically, when applied to fixed connected graphs, the proposed algorithm achieves linear convergence in solving equality-constrained DO with typical ET strategies; for UJSC graphs, it enables linear convergence in solving unconstrained DO. To further address inequality constraints, a distributed path-following ET-ZGS algorithm embedded with a finite-time max-consensus protocol is provided over UJSC digraphs, leveraging the barrier method akin to the interior-point method. Finally, two numerical examples are performed to verify the efficiency of the proposed algorithms.
Keywords:
Constrained distributed optimization
zero-gradient-sum (ZGS) method
event-triggered scheme
linear convergence
switching digraphs
Journal
I
IF:
7.9
Papers:
2.5K
Citations:
10.0K