Semistability-Based Convergence Analysis for Paracontracting Multiagent Coordination Optimization

Recently we have proposed a new class of swarm optimization algorithms called the Multiagent Coordination Optimization (MCO) [1]–[3] algorithm inspired by swarm intelligence and consensus protocols for multiagent coordination in [4]–[7]. This new algorithm is a new optimization technique based not only on swarm intelligence [8] which simulates the bio-inspired behavior, but also on cooperative control of autonomous agents. The MCO algorithm starts with a set of random solutions for agents which can communicate with each other. The agents then move through the solution space based on the evaluation of their cost functional and neighbor-toneighbor rules like multiagent consensus protocols [4]–[7], [9], [10]. Detailed convergence analysis for MCO has been conducted in the companion report [3]. In this sequential report, we first propose a paracontraction [11] based MCO algorithm and then implement the paracontracting MCO algorithm in a parallel computing way by introducing MATLAB R © built-in function parfor into the paracontracting MCO algorithm. Then we rigorously analyze the global convergence of the paracontracting MCO algorithm by means of semistability theory [4], [12]. This sequential report can be viewed as an addendum to the companion report [3].