Inertia weight strategies in GbLN-PSO for optimum solution
Particle Swarm Optimization (PSO) is the popular metaheuristic search algorithm that is inspired by the social learning of birds and fish. In the PSO algorithm, inertia weight is an important parameter to determine the searching ability of each particle. When the selected inertia weight is not suita...
Saved in:
| Main Authors: | , |
|---|---|
| Format: | Conference or Workshop Item |
| Language: | English English |
| Published: |
Institute of Electrical and Electronics Engineers Inc.
2023
|
| Subjects: | |
| Online Access: | http://umpir.ump.edu.my/id/eprint/40373/1/Inertia%20weight%20strategies%20in%20GbLN-PSO%20for%20optimum%20solution.pdf http://umpir.ump.edu.my/id/eprint/40373/2/Inertia%20weight%20strategies%20in%20GbLN-PSO%20for%20optimum%20solution_ABS.pdf http://umpir.ump.edu.my/id/eprint/40373/ https://doi.org/10.1109/ICSECS58457.2023.10256385 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| Summary: | Particle Swarm Optimization (PSO) is the popular metaheuristic search algorithm that is inspired by the social learning of birds and fish. In the PSO algorithm, inertia weight is an important parameter to determine the searching ability of each particle. When the selected inertia weight is not suitable, the searching particles are more focused on one direction or area nearest to the local best. Therefore, the movement of the particles is limited and not spreading during the search process. Thus, this will cause the particles fast to converge. As the result, the particle is trapped in local optimal. To overcome this problem, we used three different inertia weight strategies such as Constant Inertia Weight (CIW), Random Inertia Weight (RIW), and Linear Decreasing Inertia Weight (LDIW) to analyze the impact of inertia weight on the performance of Conventional PSO and the enhancement of PSO called Global Best Local Neighborhood-PSO (GbLN-PSO) algorithm. In order to test the performance of the three different inertia weight strategies, we test these algorithms in different sizes of search space with random values. Based on the comparison result of 30 simulations, it shows that GbLN-PSO using RIW was producing a better search result compared to CIW and LDIW. Furthermore, the result shows an improvement in GbLN-PSO searching ability. |
|---|