Self-adaptive particle swarm optimization: a review and analysis of convergence

Abstract

Particle swarm optimization (PSO) is a population-based, stochastic search algorithm inspired by the flocking behaviour of birds. The PSO algorithm has been shown to be rather sensitive to its control parameters and thus performance may be greatly improved by employing appropriately tuned parameters. However, parameter tuning is typically a time-intensive empirical process. Furthermore, a priori parameter tuning makes the implicit assumption that the optimal parameters of the PSO algorithm are not timedependent. To address these issues, self-adaptive particle swarm optimization (SAPSO) algorithms adapt their control parameters throughout execution. While there is a wide variety of such SAPSO algorithms in the literature, their behaviours are not well understood. Specifically, it is unknown whether these SAPSO algorithms will even exhibit convergent behaviour. This paper addresses this lack of understanding by investigating the convergence behaviours of 18 SAPSO algorithms both analytically and empirically. This paper also empirically examines whether the adapted parameters reach a stable point and whether the final parameter values adhere to a well-known convergence criterion. The results depict a grim state for SAPSO algorithms; over half of the SAPSO algorithms exhibit divergent behaviour while many others prematurely converge.

Optimal parameter regions and the time-dependence of control parameter values for the particle swarm optimization algorithm

“…To alleviate the issue of time-sensitive parameter values, various self-adaptive particle swarm optimization (SAPSO) algorithms that adapt their control parameters over time have been proposed [12,18,19,20,21,22,23]. While many such adaptive schemes have been proposed, their performance has largely been unconvincing [24,25,26,27]. However, the poor performance of SAPSO algorithms can be somewhat explained by the fact that such algorithms are concurrently optimizing two highly inter-dependent continuous search problems.…”

Swarm Intelligence-Based Optimization Techniques for Dynamic Response and Power Quality Enhancement of AC Microgrids: A Comprehensive Review

“…Several studies [117]- [119] have attempted to dynamically adjust these parameters to achieve the mentioned characteristics of the PSO for obtaining an optimal solution of the studied optimization problem. In [120], the authors have extensively reviewed the literature to address the convergence behaviors of 18 different self-adaptive PSO algorithms both empirically and analytically. However, in the majority of the PSO based MG controllers, these parameters are chosen as constant values throughout the simulation time as shown in Table 3.…”

Hunting Optimization: An New Framework for Single Objective Optimization Problems

“…In HO, a shifted cosine function is used to control the elimination rate: v r = 0.5 * cos(ite * π/maxIte) + 0.5 (4) Then adjust N ite and Q ite by…”

“…In the past decades, many population-based stochastic search methods have been proposed, such as genetic algorithm (GA) [2], differential evolution (DE) [3], particle swarm optimization (PSO) [4], ant colony optimization (ACO) [5], artificial fish (AF) [6], artificial bee colony (ABC) [7], simulated annealing (SA) [8], etc. GA includes three main evolution strategies, i.e., selection, crossover and mutation.…”