Adaptive multi mechanism integration in the crested porcupine optimizer for global optimization and engineering...

Area of Science:

• Computational Intelligence
• Optimization Algorithms
• Engineering Applications

Background:

• The Crested Porcupine Optimizer (CPO) shows promise for engineering problems but suffers from premature convergence and low population diversity.
• Existing CPO limitations hinder its effectiveness in complex, high-dimensional optimization tasks.

Purpose of the Study:

• To propose a multi-mechanism enhanced Crested Porcupine Optimizer (SDHCPO) addressing CPO's limitations.
• To improve global exploration, local exploitation, and convergence robustness of the CPO algorithm.

Main Methods:

• Implemented Sobol-Opposition-Based Learning (Sobol-OBL) for enhanced initial population distribution.
• Integrated cosine-annealing for dynamic weight adjustment to improve convergence stability.
• Incorporated DE/rand/1 and horizontal-vertical crossover strategies to prevent premature convergence and dimensional stagnation.

Main Results:

• SDHCPO significantly outperformed seven other metaheuristic algorithms on CEC2017 and CEC2022 benchmark suites.
• The enhanced algorithm achieved superior global exploration, local exploitation accuracy, and convergence robustness.
• Empirical studies on engineering design problems confirmed SDHCPO's effectiveness, yielding best-known or highly competitive solutions.

Conclusions:

• The proposed SDHCPO effectively addresses the limitations of the original CPO, particularly premature convergence and insufficient diversity.
• SDHCPO demonstrates significant potential for complex real-world engineering optimization tasks due to its enhanced performance.
• The multi-mechanism enhancement strategies contribute to the algorithm's robustness and broad applicability.