ANWOA: an adaptive nonlinear whale optimization algorithm for high-dimensional optimization problems

One of the most competitive nature-inspired metaheuristic optimization algorithms is the whale optimization algorithm (WOA). This algorithm is proven awesome in solving complex and constrained multi-objective problems. It is also popularly used as a feature selection algorithm while solving non-deterministic polynomial-time hardness (NP-hard) problems. Many enhancements have been introduced in the literature for the WOA resulting in better optimization algorithms. Differently from these research efforts, this paper presents a novel version of the WOA called ANWOA. ANWOA considers producing two types of discrete chaotic maps that have suitable period states, and the highest sensitivity to initial conditions, randomness, and stability which in turn leads to optimal initial population selection and thus global optimality. The presented ANWOA uses two nonlinear parameters instead of the two linear ones which permeate both the exploration and exploitation phases of WOA, leading to accelerated convergence, better accuracy, and influential improvement in the spiral updating position. Additionally, a dynamic inertia weight coefficient is utilized to attain a suitable balance between the exploration and exploitation phases meanwhile improving the convergence speed. Furthermore, ANWOA uses circle map values that influence each random factor in the WOA and consequently ensuring not trapped in local optima with a promoted global optimum search. The empirical analysis is conducted in thirty-three benchmark functions, and the results show that the introduced novel algorithm is the most competitive one.