A novel discrete grey wolf optimizer for solving the bounded knapsack problem

Abstract

Grey Wolf Optimizer (GWO) is a recently proposed metaheuristic optimizer inspired by the leadership hierarchy and hunting mechanism of grey wolves. In order to solve the bounded knapsack problem by the GWO, a novel Discrete Grey Wolf Optimizer (DGWO) is proposed in this paper. On the basis of DGWO, the crossover strategy of the genetic algorithm is introduced to enhance its local search ability, and infeasible solutions are processed by a Repair and Optimization method based on the greedy strategy, which could not only ensure the effectiveness but also speed up the convergence. Experiment using three kinds of large-scale instances of the bounded knapsack problem is carried out to verify the validity and stability of the DGWO. By comparing and analyzing the results with other well-established algorithms, computational results show that the convergence speed of the DGWO is faster than that of other algorithms, solutions of these instances of the bounded knapsack problem are all well obtained with approximation ratio bound close to 1.