MO-ABC/DE - Multiobjective Artificial Bee Colony with Differential Evolution for unconstrained multiobjective optimization

Abstract

Multiobjective optimization (MO) is attracting much attention of researchers in the last years. This is because most of the currently addressed optimization problems need to optimize multiple conflicting objective functions simultaneously. In this work, we propose a new algorithm called Multiobjective Artificial Bee Colony with Differential Evolution (MO-ABC/DE) for solving a set of unconstrained multiobjective optimization problems. The MO-ABC/DE algorithm is a new hybrid approach that combines the collective intelligence of the honey bee swarms (Artificial Bee Colony - ABC) with the properties of the Differential Evolution (DE). To analyse the performance of our metaheuristics we solve ten unconstrained multiobjective problems defined for the CEC 2009 Special Session on “Performance Assessment of Constrained / Bound Constrained Multiobjective Optimization Algorithms”. These test problems optimize two or three objective functions and present different properties of separability, modality, and geometry in their Pareto fronts. As we will see, the MO-ABC/DE algorithm works well on most test problems, proving to be an important tool to solve multiobjective unconstrained optimization problems.