A Novel Set-Based Particle Swarm
Optimization
Method for Discrete Optimization
Problems
Wei-Neng Chen, Student Member, IEEE, Jun Zhang, Senior Member, IEEE,
Henry S. H. Chung, Senior Member, IEEE,
Wen-Liang Zhong, Wei-Gang Wu, Member, IEEE, and Yu-hui Shi, Senior
Member, IEEE
Reporter : Yu Chih Lin
Outline
• Abstract
• Introduction
• Methodology
• Results
Abstract
• Particle swarm optimization (PSO) is
predominately used to find solutions for
continuous optimization problems.
• Proposed S-PSO features the following
characteristics
Abstract
• Discrete PSO versions based on S-PSO are
tested on two famous COPs
– The traveling salesman problem and the
multidimensional knapsack problem.
Introduction
• The algorithm is inspired by the social
interaction behavior of birds flocking and fish
schooling.
• By now, PSO has become one of the most
popular optimization techniques for solving
continuous optimization problems.
Introduction
• In order to define a more general frame for a
discrete PSO (DPSO), several approaches have
been developed in the literature.
• To obtain a feasible solution to the problem,
the algorithms need a defuzzification method
to decode the fuzzy matrix into a feasible
solution.
Methodology
• Clerc and Kennedy analyzed the convergence
behavior of PSO in detail and introduced a
constriction factor.
Methodology
• Term the global version GPSO, the local
version with URing topology ULPSO, the local
version with von Neumann topology VPSO
Results