A sterna migration algorithm-based efficient bionic engineering optimization algorithm.

Area of Science:

• Computational Intelligence
• Optimization Algorithms
• Metaheuristic Computing

Background:

• Existing metaheuristic algorithms often struggle to balance global exploration and local exploitation effectively.
• There is a continuous need for novel optimization techniques that exhibit superior performance on complex, high-dimensional problems.

Purpose of the Study:

• To introduce the Sterna Migration Algorithm (StMA), a new metaheuristic optimization method.
• To evaluate StMA's performance against established algorithms on benchmark functions and engineering design problems.
• To demonstrate StMA's capability in achieving a dynamic balance between exploration and exploitation for improved optimization.

Main Methods:

• Developed StMA integrating multi-cluster sectoral diffusion, leader-follower dynamics, adaptive perturbation, and multi-phase termination.
• Systematically evaluated StMA on CEC2023 and CEC2014 benchmark functions and constrained engineering design problems.
• Conducted comparative analysis against mainstream population-based algorithms over 30 independent runs per problem.

Main Results:

• StMA significantly outperformed competitors on 23 of 30 CEC2014 functions, showing 100% superiority on unimodal functions.
• Demonstrated improved convergence efficiency (37.2% decrease in average generations) and solution accuracy (14.7%-92.3% error reduction).
• Achieved best overall performance on six constrained engineering design problems, validating robustness and adaptability.

Conclusions:

• StMA offers a novel approach to metaheuristic optimization, effectively balancing exploration and exploitation.
• The algorithm exhibits superior performance, efficiency, and robustness on diverse and complex optimization tasks.
• StMA provides a strong foundation for developing next-generation metaheuristic algorithms for challenging scenarios.