An Efficient Parameter Optimization Approach Based on Real-Coded Genetic Algorithm

Abstract

In this paper, we present an efficient parameter optimization approach based on real-coded genetic algorithm, which breeds selectively offspring based on difference between individuals. A new crossover operator is proposed which generates offspring around the center of mass of parents with Laplace distribution. A set of 14 test problems available in the global optimization literature is used to evaluate the performance of proposed genetic algorithm. The comparative study shows that the proposed genetic algorithm performs quite well for parameter optimization problem.