Grey wolf optimizer with an enhanced hierarchy and its application to the wireless sensor network coverage optimization problem

Abstract Grey wolf optimizer (GWO), which is inspired by the social behaviours of grey wolf packs, is a nature-inspired and population-based algorithm. The GWO technique has the advantage of conceptual simplicity and shows good results for solving various real-world problems. However, this technique has the drawback of premature convergence and is prone to stagnation in local optima. The leadership hierarchy is the paramount characteristic of the GWO and influences its searching precision. Therefore, a grey wolf optimizer with enhanced hierarchy (GWO-EH) is proposed to overcome these deficiencies. Firstly, fitness-based self-adaptive weight coefficients are introduced to better imitate the hierarchy of the grey wolves, which also have a positive effect on the convergence speed. Then, we propose an improved position-updating equation to enhance the leadership of the high-ranking wolves, whereby the global exploration ability of the GWO is strengthened. Finally, the strategy of repositioning wolves around the leading wolves is designed to keep a perfect balance between exploration and exploitation. The search ability of the GWO-EH is thoroughly compared with the GWO itself, some promising GWO variants, and several well-established algorithms on twenty-three widely used benchmark functions. Empirical studies reveal that GWO-EH has a competitive overall performance according to the average value (standard deviation), Wilcoxon rank-sum test results, and convergence curve. Moreover, our method is applied to address the wireless sensor network coverage optimization problem, and the applicability and validity of the GWO-EH are indicated by the experimental results.