Effects of Random Values for Particle Swarm Optimization Algorithm
Particle swarm optimization (PSO) algorithm is generally improved by adaptively adjusting the inertia weight or combining with other evolution algorithms. However, in most modified PSO algorithms, the random values are always generated by uniform distribution in the range of [0, 1]. In this study, t...
| Main Authors: | , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
MDPI AG
2018-02-01
|
| Series: | Algorithms |
| Subjects: | |
| Online Access: | http://www.mdpi.com/1999-4893/11/2/23 |
| _version_ | 1818363615387844608 |
|---|---|
| author | Hou-Ping Dai Dong-Dong Chen Zhou-Shun Zheng |
| author_facet | Hou-Ping Dai Dong-Dong Chen Zhou-Shun Zheng |
| author_sort | Hou-Ping Dai |
| collection | DOAJ |
| description | Particle swarm optimization (PSO) algorithm is generally improved by adaptively adjusting the inertia weight or combining with other evolution algorithms. However, in most modified PSO algorithms, the random values are always generated by uniform distribution in the range of [0, 1]. In this study, the random values, which are generated by uniform distribution in the ranges of [0, 1] and [−1, 1], and Gauss distribution with mean 0 and variance 1 (
U
[
0
,
1
]
,
U
[
−
1
,
1
]
and
G
(
0
,
1
)
), are respectively used in the standard PSO and linear decreasing inertia weight (LDIW) PSO algorithms. For comparison, the deterministic PSO algorithm, in which the random values are set as 0.5, is also investigated in this study. Some benchmark functions and the pressure vessel design problem are selected to test these algorithms with different types of random values in three space dimensions (10, 30, and 100). The experimental results show that the standard PSO and LDIW-PSO algorithms with random values generated by
U
[
−
1
,
1
]
or
G
(
0
,
1
)
are more likely to avoid falling into local optima and quickly obtain the global optima. This is because the large-scale random values can expand the range of particle velocity to make the particle more likely to escape from local optima and obtain the global optima. Although the random values generated by
U
[
−
1
,
1
]
or
G
(
0
,
1
)
are beneficial to improve the global searching ability, the local searching ability for a low dimensional practical optimization problem may be decreased due to the finite particles. |
| first_indexed | 2024-12-13T21:51:17Z |
| format | Article |
| id | doaj.art-8a483417fcaf49c58735946077b7a97d |
| institution | Directory Open Access Journal |
| issn | 1999-4893 |
| language | English |
| last_indexed | 2024-12-13T21:51:17Z |
| publishDate | 2018-02-01 |
| publisher | MDPI AG |
| record_format | Article |
| series | Algorithms |
| spelling | doaj.art-8a483417fcaf49c58735946077b7a97d2022-12-21T23:30:16ZengMDPI AGAlgorithms1999-48932018-02-011122310.3390/a11020023a11020023Effects of Random Values for Particle Swarm Optimization AlgorithmHou-Ping Dai0Dong-Dong Chen1Zhou-Shun Zheng2School of Mathematics and Statistics, Central South University, Changsha 410083, ChinaSchool of Mathematics and Statistics, Central South University, Changsha 410083, ChinaSchool of Mathematics and Statistics, Central South University, Changsha 410083, ChinaParticle swarm optimization (PSO) algorithm is generally improved by adaptively adjusting the inertia weight or combining with other evolution algorithms. However, in most modified PSO algorithms, the random values are always generated by uniform distribution in the range of [0, 1]. In this study, the random values, which are generated by uniform distribution in the ranges of [0, 1] and [−1, 1], and Gauss distribution with mean 0 and variance 1 ( U [ 0 , 1 ] , U [ − 1 , 1 ] and G ( 0 , 1 ) ), are respectively used in the standard PSO and linear decreasing inertia weight (LDIW) PSO algorithms. For comparison, the deterministic PSO algorithm, in which the random values are set as 0.5, is also investigated in this study. Some benchmark functions and the pressure vessel design problem are selected to test these algorithms with different types of random values in three space dimensions (10, 30, and 100). The experimental results show that the standard PSO and LDIW-PSO algorithms with random values generated by U [ − 1 , 1 ] or G ( 0 , 1 ) are more likely to avoid falling into local optima and quickly obtain the global optima. This is because the large-scale random values can expand the range of particle velocity to make the particle more likely to escape from local optima and obtain the global optima. Although the random values generated by U [ − 1 , 1 ] or G ( 0 , 1 ) are beneficial to improve the global searching ability, the local searching ability for a low dimensional practical optimization problem may be decreased due to the finite particles.http://www.mdpi.com/1999-4893/11/2/23particle swarm optimization algorithmrandom valuesuniform distributiongauss distribution |
| spellingShingle | Hou-Ping Dai Dong-Dong Chen Zhou-Shun Zheng Effects of Random Values for Particle Swarm Optimization Algorithm Algorithms particle swarm optimization algorithm random values uniform distribution gauss distribution |
| title | Effects of Random Values for Particle Swarm Optimization Algorithm |
| title_full | Effects of Random Values for Particle Swarm Optimization Algorithm |
| title_fullStr | Effects of Random Values for Particle Swarm Optimization Algorithm |
| title_full_unstemmed | Effects of Random Values for Particle Swarm Optimization Algorithm |
| title_short | Effects of Random Values for Particle Swarm Optimization Algorithm |
| title_sort | effects of random values for particle swarm optimization algorithm |
| topic | particle swarm optimization algorithm random values uniform distribution gauss distribution |
| url | http://www.mdpi.com/1999-4893/11/2/23 |
| work_keys_str_mv | AT houpingdai effectsofrandomvaluesforparticleswarmoptimizationalgorithm AT dongdongchen effectsofrandomvaluesforparticleswarmoptimizationalgorithm AT zhoushunzheng effectsofrandomvaluesforparticleswarmoptimizationalgorithm |