An improved hippopotamus optimization algorithm based on adaptive development and solution diversity enhancement.

Area of Science:

• Computational Intelligence
• Optimization Algorithms
• Metaheuristics

Background:

• Traditional hippopotamus optimization algorithm faces limitations in convergence and solution diversity for complex problems.
• High-dimensional and multimodal functions pose significant challenges for existing optimization techniques.

Purpose of the Study:

• To propose an improved hippopotamus optimization algorithm addressing limitations in convergence and solution diversity.
• To enhance the global search capability and avoid local optima in complex optimization tasks.

Main Methods:

• Incorporated chaotic map initialization for optimized initial population distribution.
• Introduced an adaptive exploitation mechanism to balance exploration and exploitation phases.
• Implemented a solution diversity enhancement strategy using nonlinear perturbations.

Main Results:

• The improved algorithm demonstrated significantly better convergence speed and solution accuracy compared to the original algorithm.
• Outperformed other mainstream optimization algorithms on standard benchmark functions (CEC17, CEC22).
• Statistical analysis confirmed superior exploration-exploitation balance, especially for high-dimensional multimodal functions.

Conclusions:

• The enhanced hippopotamus optimization algorithm offers improved performance for complex, high-dimensional problems.
• The study provides effective strategies for advancing metaheuristic optimization techniques.
• The proposed enhancements offer valuable insights for future research in computational intelligence.