An new crossover operator for real-coded genetic algorithm with selective breeding based on difference between individuals

In this paper, we present an efficient crossover operator for real-coded genetic algorithm that breeds offspring based on difference between individuals. In the proposed crossover operator, offspring are generated following a promising direction with Laplace distribution based on the center of mass of parents. A set of 15 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 and outperforms other algorithms.