Multi-Objective Harris Hawks Optimizer For Multiobjective Optimization Problems

Abstract

In this paper, a multi-objective version of the Harris Hawk Optimizer algorithm (HHO) is proposed, which is called Multi-Objective Harris Hawk Optimization (MOHHO). In the MOHHO algorithm, preserving the structure of the HHO algorithm, an archive repository has been added to the HHO algorithm to save and retrieve the Pareto optimal results. This repository is used for simulating the positions and solutions of the hawks. The archive member in the least populated area from this archive is selected using the roulette wheel process. This archive member is utilized as the rabbit in the proposed MOHHO algorithm. To show the performance of the MOHHO algorithm, we have taken unconstrained test functions known as ZDT from the literature. For the multi objective benchmarks, the MOHHO algorithm was compared with MOALO (Multi-objective AntLion optimizer) and MODA (Multi-objective Dragonfly optimizer) algorithms. Inverted Generational Distance (IGD) metric was used for ZDT benchmark comparison studies. The comparison results show that the proposed algorithm gives better results than the MOALO and MODA algorithms in terms of IGD metric for all test functions.