Effects of memory and genetic operators on Artificial Bee Colony algorithm for Single Container Loading problem

Abstract

The Artificial Bee Colony (ABC) algorithm is widely used to achieve optimum solution in a short time in integer-based optimization problems. However, the complexity of integer-based problems such as Knapsack Problems (KP) requires robust algorithms to avoid excessive solution search time. ABC algorithm that provides both the exploitation and the exploration approach is used as an alternative approach for various KP problems in the literature. However, it is rarely used for the Single Container Loading problem (SCLP) which is an important part of the transportation systems. In this study, the exploitation and exploration aspects of the ABC algorithm are improved by using memory mechanisms and genetic operators to develop three different hybrid ABC algorithms. The developed algorithms and the basic ABC algorithm are applied to a SCLP dataset from the literature to observe the effects of the memory mechanism and the genetic operators separately. Besides, a joint hybrid ABC algorithm using both reinforcement approaches is proposed to solve the SCLP. The results show that the joint hybrid ABC algorithm has emerged as a promising approach to solving SCLP with an average performance, and the genetic operators are more effective than the memory mechanism to develop a hybrid ABC algorithm.