.png){kind=icon}
.png){kind=icon}
Return
A novel improved particle swarm optimization algorithm based on individual difference evolution
.png){kind=icon}
PRE
AI
DOI：10.1016/j.asoc.2017.04.025.png)
.png){kind=icon}
Original request for help.png){kind=icon}
Share.png){kind=icon}
Save
Abstract
En 中文
As a well-known stochastic optimization algorithm, the particle swarm optimization (PSO) algorithm has attracted the attention of many researchers all over the world, which has resulted in many variants of the basic algorithm, in addition to a vast number of parameter selection/control strategies. However, most of these algorithms evolve their population using a single fixed pattern, thereby reducing the intelligence of the entire swarm. Some PSO-variants adopt a multimode evolutionary strategy, but lack dynamic adaptability. Furthermore, competition among particles is ignored, with no consideration of individual thinking or decision-making ability. This paper introduces an evolution mechanism based on individual difference, and proposes a novel improved PSO algorithm based on individual difference evolution (IDE-PSO). This algorithm allocates a competition coefficient called the emotional status to each particle for quantifying individual differences, separates the entire swarm into three subgroups, and selects the specific evolutionary method for each particle according to its emotional status and current fitness. The value of the coefficient is adjusted dynamically according to the evolutionary performance of each particle. A modified restarting strategy is employed to regenerate corresponding particles and enhance the diversity of the population. For a series of benchmark functions, simulation results show the effectiveness of the proposed IDE-PSO, which outperforms many state-of-the-art evolutionary algorithms in terms of convergence, robustness, and scalability. (C) 2017 Elsevier B.V. All rights reserved.
Keywords:
Particle swarm optimization
Individual difference
Dynamic adjustment
Subgroup
motional PSO
Psychology model
AI Summary
Key information extracted from the uploaded paper, including a brief overview, abstract, background, key highlights, visual analysis, and future outlook.