Periodicals of Engineering and Natural Sciences ISSN 2303-4521 Vol. 9, No. 2, March 2021, pp.400-409 1 A hybrid Grey Wolf optimizer with multi-population differential evolution for global optimization problems Nuha Sami Mohsin 1 , Buthainah F. Abd 2 , Rafah Shihab Alhamadani 3 1 Faculty of Education Ibn Rushd for Human Sciences, Baghdad University, Baghdad, Iraq 2 Smart Cities Dept, University of Information Technology and Communications, Baghdad, Iraq 3 Iraqi Commission for Computers & Informatics, Informatics Institute for Postgraduate Studies, Baghdad, Iraq ABSTRACT The optimization field is the process of solving an optimization problem using an optimization algorithm. Therefore, studying this research field requires to study both of optimization problems and algorithms. In this paper, a hybrid optimization algorithm based on differential evolution (DE) and grey wolf optimizer (GWO) is proposed. The proposed algorithm which is called “MDE-GWONM” is better than the original versions in terms of the balancing between exploration and exploitation. The results of implementing MDE- GWONM over nine benchmark test functions showed the performance is superior as compared to other stat of arts optimization algorithms Keywords: Optimization, Metahuristics , Grey Wolf Optimizater, Differential Evolution, Multi- Population Corresponding Author: Nuha Sami Mohsin, Faculty of Education Ibn Rushd for Human Sciences, Baghdad University, Baghdad, Iraq, Email: nuha.sami@ircoedu.uobaghdad.edu.iq 1. Introduction Optimization problems are problems that form a unique class of problems while trying to either minimize or maximize the mathematical function of several variables with respect to certain constraints. This general framework can be used to model several problems, both theoretical and real-world. The structure of mathematical models (or mathematical programming model) can generally be represented as follows [1]–[3]:    (  ), ( = 1, 2, 3, … , ) (1)   ℎ  (  ) = 0, ( = 1, 2, 3, … , ),   (  ) ≤ 0, ( = 1, 2, 3, … , ) where  is the decision variables, and   (), ℎ  (), and   () are functions of the design vector. = (  1 , 2 , 3 ,…,  )  (2) Metaheuristics have been commonly used in the field of optimization compared to other methods due to the simplicity and robustness of their outcome when used in several fields. Several studies have been conducted in the area of metaheuristics, including the introduction of novel methods, performance evaluations and applications [4]–[6]. Meanwhile, it is still believed that the field of metaheuristics is yet to mature compared to mathematics, physics, or chemistry [7]. With time, several studies are anticipated on the issues facing metaheuristic computing.