Enhancing Population Diversity for Genetic Algorithms

The premature convergence may lead the genetic algorithms (GAs) to a local optimum but not a global one. Maintaining the population diversity in GAs, or minimize its loss, may alleviate this problem to a certain extent. A novel selector based on eugenics (EBSelector) has been proposed to faciliate effective selection of individuals to perform crossover, inspired by the eugenic theory about how to make familial disease less happen and to produce high-quality offspring. Demonstrated through a suite of benchmark test functions, the proposed algorithm is shown competitive performance with improved convergence speed.