TY - GEN
T1 - An Improved Competitive Mechanism based Particle Swarm optimization Algorithm for Multi-Objective optimization
AU - Yuen, Man Chung
AU - Ng, Sin Chun
AU - Leung, Man Fai
N1 - Publisher Copyright:
© 2020 IEEE.
PY - 2020/9
Y1 - 2020/9
N2 - In this paper, an improved Competitive Mechanism-based Particle Swarm optimization algorithm called MCMOPSO is presented for multi-objective optimization. The algorithm consists of two main contributions: a new leader selection and the analysis of inertia weight. The new multi competition leader selection is introduced which is based on the pairwise competition. It will not only guide the particles to fly to the winner by comparing the nearest angle for two randomly selected elite particles, but also lead the particles to fly to the winner by comparing the nearest angle or farthest angle for several randomly selected elite particles in each iteration. To strike a balance between the exploration and exploitation of the velocity update equation for the original competitive mechanism-based MOPSO algorithm (CMOPSO), the influence of various inertia weights is investigated to control the previous velocity of each particle. The simulation results show that the proposed algorithm is outperformed four other famous multi-objective particle swarm optimization algorithms in thirty-seven benchmark test problems in terms of inverted generational distance.
AB - In this paper, an improved Competitive Mechanism-based Particle Swarm optimization algorithm called MCMOPSO is presented for multi-objective optimization. The algorithm consists of two main contributions: a new leader selection and the analysis of inertia weight. The new multi competition leader selection is introduced which is based on the pairwise competition. It will not only guide the particles to fly to the winner by comparing the nearest angle for two randomly selected elite particles, but also lead the particles to fly to the winner by comparing the nearest angle or farthest angle for several randomly selected elite particles in each iteration. To strike a balance between the exploration and exploitation of the velocity update equation for the original competitive mechanism-based MOPSO algorithm (CMOPSO), the influence of various inertia weights is investigated to control the previous velocity of each particle. The simulation results show that the proposed algorithm is outperformed four other famous multi-objective particle swarm optimization algorithms in thirty-seven benchmark test problems in terms of inverted generational distance.
KW - competitive mechanism
KW - multi-objective optimization problem
KW - particle swarm optimization
UR - http://www.scopus.com/inward/record.url?scp=85093973260&partnerID=8YFLogxK
U2 - 10.1109/ICIST49303.2020.9202228
DO - 10.1109/ICIST49303.2020.9202228
M3 - Conference contribution
AN - SCOPUS:85093973260
T3 - 10th International Conference on Information Science and Technology, ICIST 2020
SP - 209
EP - 218
BT - 10th International Conference on Information Science and Technology, ICIST 2020
T2 - 10th International Conference on Information Science and Technology, ICIST 2020
Y2 - 9 September 2020 through 15 September 2020
ER -