An enhancement of selection and crossover operations in real-coded genetic algorithm for large-dimensionality optimization

作者：Kwak Noh Sung; Lee Jongsoo^*

来源：Journal of Mechanical Science and Technology, 2016, 30(1): 237-247.

DOI：10.1007/s12206-015-1227-2

• Real-coded genetic algorithm
• Large-scale global optimization
• Subpopulation
• Probabilistic crossover

The present study aims to implement a new selection method and a novel crossover operation in a real-coded genetic algorithm. The proposed selection method facilitates the establishment of a successively evolved population by combining several subpopulations: an elitist subpopulation, an off-spring subpopulation and a mutated subpopulation. A probabilistic crossover is performed based on the measure of probabilistic distance between the individuals. The concept of 'allowance' is suggested to describe the level of variance in the crossover operation. A number of nonlinear/non-convex functions and engineering optimization problems are explored to verify the capacities of the proposed strategies. The results are compared with those obtained from other genetic and nature-inspired algorithms.

• 出版日期2016-1