Parameters for a genetic algorithm: An application for the order batching problem

Abstract

This article aims to validate the parameters of a genetic algorithm for the order batching problem (OBP) in warehouses by defining the parameter values offering the best solution performance. Thus, a description of the OBP and the solution approaches based on item-oriented and group-oriented genetic algorithms are introduced. Then, the characteristics of a group-oriented genetic algorithm are shown, and experiments are performed to establish the parameter values related to population size, crossover rate, elitism rate, and mutation rate. Therefore, we provide the set of parameter values for the genetic algorithm offering better quality results in terms of total distance traveled, and some recommendations to reduce the computing time of the algorithm are presented.