Modular Harris Hawks optimization with trend-guided differential evolution and Gaussian exploration for global...

Area of Science:

• Computational Intelligence
• Optimization Algorithms
• Metaheuristics

Background:

• High-dimensional and complex optimization tasks often suffer from stagnation and instability.
• Existing metaheuristics, including Harris Hawks Optimization (HHO), face challenges in maintaining diversity and convergence.
• Novel algorithmic approaches are needed to address these limitations.

Purpose of the Study:

• To present DEHHO, a novel, modular, and lightweight variant of Harris Hawks Optimization (HHO).
• To enhance optimization performance in high-dimensional and complex problems by mitigating stagnation and directional instability.
• To validate the effectiveness of DEHHO against state-of-the-art algorithms.

Main Methods:

• Integration of a Gaussian-based stochastic perturbation mechanism for exploration diversity.
• Incorporation of a Trend-Guided Differential Evolution (DE) update with a momentum vector for exploitation stability.
• A dynamic balancing scheme to coordinate search phases efficiently.

Main Results:

• DEHHO statistically outperformed 10 peer algorithms on CEC 2017 and CEC 2020 benchmark suites (up to 100 dimensions).
• Demonstrated superior convergence accuracy, robustness, and scalability compared to advanced HHO variants and mainstream metaheuristics.
• Ablation studies confirmed the efficacy and complementarity of the proposed mechanisms.

Conclusions:

• DEHHO is a reliable and effective solver for complex numerical and engineering design optimization problems.
• The synergistic integration of exploration and exploitation strategies significantly enhances optimization performance.
• The proposed framework offers a robust solution for tackling challenging optimization landscapes.