A New and Efficient Genetic Algorithm with Promotion Selection Operator

Abstract

Genetic algorithm (GA) is a widely used probabilistic search optimization algorithm. In the GA, selection is an important operator to guarantee the quality of solution. Therefore, the behavior of selection operator makes a great effect on the performance of the algorithm. This paper designs a new and efficient selection operator for GA base on the idea of promotion competition. This operator simulates the rule and process of promotion competition to protect the well perform chromosomes and eliminates poor chromosomes. This is a fundamental but significant research issue in GA that may be adopted into any existing GA variants to replace any other selection operators. We design four types of experiments to comprehensively verify the behavior of the proposed promotion selection operator, by comparing it with five other existing and commonly used selection operators. The results show that promotion selection operator has a general good performance in enhancing GA in terms of solution quality, convergence speed, and running time.