A novel micro-population immune multiobjective optimization algorithm

  • Authors:

  • Qiuzhen Lin;Jianyong Chen

  • Affiliations:

  • Department of Electronic Engineering, City University of Hong Kong, 83 Tat Chee Avenue, Kowloon Tong, Hong Kong;Department of Computer Science and Technology, Shenzhen University, Shenzhen 518060, China

  • Venue:
  • Computers and Operations Research
  • Year:
  • 2013

Quantified Score

Hi-index 0.01

Visualization

Abstract

In this paper, we present a novel immune multiobjective optimization algorithm based on micro-population, which adopts a novel adaptive mutation operator for local search and an efficient fine-grained selection operator for archive update. With the external archive for storing nondominated individuals, the population diversity can be well preserved using an efficient fine-grained selection procedure performed on the micro-population. The adaptive mutation operator is executed according to the fitness values, which promotes to use relatively large steps for boundary and less-crowded individuals in high probability. Therefore, the exploratory capabilities are enhanced. When comparing the proposed algorithm with a recently proposed immune multiobjective algorithm and a scatter search multiobjective algorithm in various benchmark functions, simulations show that the proposed algorithm not only improves convergence ability but also preserves population diversity adequately in most cases.