Improved Harris's Hawk Multi-objective Optimizer Using Two-steps Initial Population Generation Method

Abstract

The population of hawks in the Harris's hawk multi-objective optimizer (HHMO) algorithm is generated using uniform distribution random number. This method does not guarantee that the solutions can be evenly distributed in the search space of the problem, which may affect the efficiency of the algorithm. Therefore, to improve the performance of HHMO algorithm, two-steps initial population generation method is proposed. This method is developed based on R-sequence and partial opposition-based learning, which is employed to generate an initial population of hawks, with the aim to achieve better initial population. Thus better convergence toward Pareto front will be obtained. The performance of the proposed improved HHMO algorithm is evaluated using a set of well-known multi-objective optimization problems. The results of numerical simulation experiment demonstrate the effectiveness of the proposed two-step initial population generation method and showed superiority of the improved HHMO algorithm compares to the HHMO. The improved HHMO can be used to improve the convergence towards the true Pareto frontier.