Modified Particle Swarm Optimization with Chaos-Based Particle Initialization and Logarithmic Decreasing Inertia Weight

• Main Authors: Murinto, Murinto, Harjoko, Agus, Hartati, Sri, Danoedoro, Projo
• Format: Article PeerReviewed
• Language: English
• Published: ICIC International 2022
• Subjects: Information Systems not elsewhere classified; Information System
• Online Access: https://repository.ugm.ac.id/282147/1/Murinto%20et%20al%20-%202022%20-%20MODIFIED%20PARTICLE%20SWARM%20OPTIMIZATION.pdf; https://repository.ugm.ac.id/282147/; http://www.icicel.org/ell/contents/2022/2/el-16-02-11.pdf

The global optimization problem can be solved using one of the algorithms, namely particle swarm optimization (PSO). The PSO algorithm is a population optimization based on swarm intelligence, which has been widely studied and is widely applied to various problems. However, PSO is often trapped in local optimal and premature convergence on complex multimodal function problems. To solve this problem, a variant of particle swarm optimization involves the chaos maps mechanism strategy and the inertia weight of standard particle swarm optimization. Chaos map is used to produce uniform particle distribution to improve the quality of the initial position of the particles. While the inertia weight used here is logarithmic decreasing inertia weight (LogDIW) to help the algorithm get out of the local optimal and make the particles continue to search in other areas of the solution space. Extensive experiments on six well-known benchmark functions with different dimensions show that the proposed PSO is superior or very competitive to several other PSO variants in dealing with complex multimodal problems. © 2022 ICIC International. All rights reserved.