TY - GEN
T1 - Identifying Species for Particle Swarm Optimization under Dynamic Environments
AU - Luo, Wenjian
AU - Sun, Juan
AU - Bu, Chenyang
AU - Yi, Ruikang
N1 - Publisher Copyright:
© 2018 IEEE.
PY - 2018/7/2
Y1 - 2018/7/2
N2 - The species techniques have been widely used in multimodal optimization, but not paid enough attention in dynamic optimization. In this paper, a novel Nearest-Better Clustering (NBC) method, called psfNBC, is proposed, which is used to identify the species for Particle Swarmoptimization (PSO) under dynamic environments. The proposed psfNBC includes three novel properties. First, we propose the strategy p to cope with the effects of outliers. Second, the strategy s is introduced to deal with the 'long-tail phenomenon Third, a random scale factor, which is named as the strategy f, is adopted. In order to evaluate the performance of the psfNBC, a framework of species-based particle swarm optimization for Dynamic optimization Problems (DOPs) is given. Within this framework, the proposed psfNBC is compared with the basic NBC as well as different combinations of the proposed strategies. The experimental results on the moving peak benchmark problems show that psfNBC has better performance.
AB - The species techniques have been widely used in multimodal optimization, but not paid enough attention in dynamic optimization. In this paper, a novel Nearest-Better Clustering (NBC) method, called psfNBC, is proposed, which is used to identify the species for Particle Swarmoptimization (PSO) under dynamic environments. The proposed psfNBC includes three novel properties. First, we propose the strategy p to cope with the effects of outliers. Second, the strategy s is introduced to deal with the 'long-tail phenomenon Third, a random scale factor, which is named as the strategy f, is adopted. In order to evaluate the performance of the psfNBC, a framework of species-based particle swarm optimization for Dynamic optimization Problems (DOPs) is given. Within this framework, the proposed psfNBC is compared with the basic NBC as well as different combinations of the proposed strategies. The experimental results on the moving peak benchmark problems show that psfNBC has better performance.
KW - dynamic optimization
KW - nearest-better clustering
KW - particle swarm optimization
KW - species identification
UR - https://www.scopus.com/pages/publications/85062772613
U2 - 10.1109/SSCI.2018.8628900
DO - 10.1109/SSCI.2018.8628900
M3 - 会议稿件
AN - SCOPUS:85062772613
T3 - Proceedings of the 2018 IEEE Symposium Series on Computational Intelligence, SSCI 2018
SP - 1921
EP - 1928
BT - Proceedings of the 2018 IEEE Symposium Series on Computational Intelligence, SSCI 2018
A2 - Sundaram, Suresh
PB - Institute of Electrical and Electronics Engineers Inc.
T2 - 8th IEEE Symposium Series on Computational Intelligence, SSCI 2018
Y2 - 18 November 2018 through 21 November 2018
ER -