Non-dominated Sorting Harris’s Hawk Multi-Objective Optimizer based on the Flush-and-Ambush Tactic

Abstract

In this paper, a new population update strategy is proposed to overcome the limitations of the non-dominated sorting Harris’s hawk multi-objective optimizer (NDSHHMO) algorithm. In the NDSHHMO algorithm, the population of hawks is updated based on the average positions of the first three best solutions in the search space. This update strategy leads to the algorithm falling into local optima due to population diversity loss, which causes poor convergence toward the true Pareto front. The proposed population update strategy is inspired by the flush-and-ambush (FA) tactic employed by the Harris’s hawks in nature. The proposed algorithm is called non-dominated sorting Harris’s hawks’ multi-objective optimizer based on the flush-and-ambush tactic (FA-NDSHHMO). The population update strategy in the FA-NDSHHMO includes two main stages, namely, updating the position of hawks using proposed flush-and-ambush movement strategy and selecting the best hawks by using a non-dominated sorting approach to be used in the next generation. The proposed population update strategy aims to improve the search ability of the algorithm, in terms of the diversity of a non-dominated solution and convergence toward the Pareto front. To evaluate the performance of the FA-NDSHHMO algorithm, a set of 10 multi-objective optimization problems has been used. The obtained results show that the new population update strategy has improved the search ability of the FA-NDSHHMO. Furthermore, the results show superiority of the FA-NDSHHMO algorithm compared to the NDSHHMO, multi-objective grasshopper and grey wolf optimization algorithms.