Incorporating a Genetic Algorithm to improve the performance of Variable Neighborhood Search

Abstract

Variable Neighborhood Search (VNS) is an efficient metaheuristics in solving optimization problems. Although VNS has been successfully applied on various problem domains, it suffers from its inefficient search exploration. To improve this limitation, VNS can be joined with a population-based search to benefit from its search exploration. In this article, a Memetic Algorithm (MA) is proposed which is based on a Genetic Algorithm (GA) incorporating VNS as a local search method. To evaluate the proposed method, it has been applied on the classical Job Shop Scheduling Problem (JSSP) as a well-known optimization problem. The experimental results show that the proposed MA outperforms the VNS method. Furthermore, compared to the state-of-the-art Evolutionary Algorithms (EAs) proposed to solve JSSP, the proposed method offers competitive solutions.