Integrating Differential Evolution into Gazelle Optimization for advanced global optimization and engineering applications

Abstract

The Gazelle Optimization Algorithm (GOA) is an innovative metaheuristic inspired by the survival tactics of gazelles in predator-rich environments. While GOA demonstrates notable advantages in solving unimodal, multimodal, and engineering optimization problems, it struggles with local optima and slow convergence in high-dimensional and non-convex scenarios. This paper proposes the Hybrid Gazelle Optimization Algorithm with Differential Evolution (HGOADE), which combines Differential Evolution (DE) with GOA to leverage their complementary strengths for addressing limitations. HGOADE initializes a population of candidate solutions using GOA, then enhances these solutions through DE’s mutation and crossover operations. The algorithm subsequently employs GOA’s exploration and exploitation phases to refine the solutions. By integrating DE’s robust exploration capabilities with GOA’s dynamic search patterns, HGOADE aims to improve global and local search performance. The effectiveness of HGOADE is validated through experiments on benchmark functions from the CEC 2017, CEC 2020, CEC 2022 suite, comparing with ten established optimization techniques, including classical GOA, Salp Swarm Algorithm (SSA), Grey Wolf Optimizer (GWO), Gravitational Search Algorithm (GSA), Arithmetic Optimization Algorithm (AOA), Constriction Coefficient-Based Particle Swarm Optimization Gravitational Search Algorithm (CPSOGSA), Sine Cosine Algorithm (SCA), Particle Swarm Optimization (PSO), Biogeography-Based Optimization (BBO), and DE. Additionally, the performance of HGOADE is assessed against prominent winners from CEC competitions, specifically CMA-ES, LSHADEcnEpSin, and LSHADESPACMA, using the CEC-2017 test suite. Statistical analyses using the Wilcoxon Rank Sum Test and Wilcoxon Signed-Rank Test, along with the Weighted Aggregated Sum Product Assessment (WASPAS) method, confirm that HGOADE significantly outperforms existing algorithms in terms of solution quality and convergence speed. HGOADE’s superiority is validated through six complex engineering design problems, demonstrating its higher feasibility and effectiveness than GOA and other methods. This paper addresses GOA’s shortcomings and advances metaheuristic optimization by integrating DE strategies, offering valuable insights and practical applications for global optimization and engineering design.