Enhancing Pelican Optimization Algorithm with Differential Evolution: A Novel Hybrid Metaheuristic Approach

Abstract

In the field of solutions for composite objective functions, the problem of identifying a proper trade-off between exploitation and exploration is still urgent. Classical methods can hardly avoid early iteration convergence or be insufficient in terms of searching throughout the space of potential solutions, especially when dealing with multi-variate multi-dimensional problems. To overcome this problem, this work proposes a combination of the pelican optimization algorithm (POA) and differential evolution (DE), known as the POA-DE metaheuristic method, which comprises the explorative characteristic of POA and the exploitative feature of DE. The main issue dealt with in this work relates to the conflict of global search and local exploitation in the context of solving complex optimization tasks. In global exploration, the POA technique is applied to improve the performances of the search in the large area, and the DE method is used in the local search space for improving the solution. To this end, the proposed solution hybrid model tries to avoid the shortcomings associated with using either of the two key aspects when used independently. To support the results obtained through POA-DE, it is necessary to perform the intensive empirical examination of several benchmark functions. The results also show that the proposed method has achieved better stability, efficiency, and convergence speed than the basic POA. Therefore, extending the hybrid optimization techniques is significant in enhancing the meta-heuristic algorithms that form a powerful tool to solve the optimization problems.