Event-Triggered Consensus-Based Optimization Algorithm for Smooth and Strongly Convex Cost Functions
Naoki Hayashi, Tomohiro Sugiura, Yuichi Kajiyama, Shigemasa Takai · 2018
In this paper, we propose a distributed event-triggered algorithm for separable optimization problems with smooth and strongly convex cost functions. We consider a multiagent system where each agent has a state and an auxiliary variable for the estimates of the optimal solution and the average gradient of the entire cost function. Agents exchange their states and auxiliary variables when the differences at the current time and the last trigger time exceed a threshold. We show that the proposed event-triggered algorithm with an exponentially decaying trigger condition linearly converges to the optimal solution.