Gorilla optimization algorithm combining sine cosine and cauchy variations and its engineering applications.

Area of Science:

• Computational Intelligence
• Optimization Algorithms
• Metaheuristics

Background:

• The artificial gorilla troops optimizer algorithm (AGTO) suffers from limited exploration ability, population diversity loss, and premature convergence.
• These limitations hinder effective optimization, particularly in later stages of search.

Purpose of the Study:

• To enhance the AGTO algorithm by integrating positive cosine and Cauchy variance strategies.
• To improve convergence speed, accuracy, and global optimization capabilities.

Main Methods:

• Introduced refractive reverse learning for population initialization to increase diversity.
• Incorporated a positive cosine strategy and adaptive weight factors for balanced exploration and exploitation.
• Applied Cauchy variance to follower position updates for improved global search.

Main Results:

• SCAGTO demonstrated significant improvements in convergence speed and accuracy on 30 classical test functions.
• The algorithm showed robust performance and advantages in solving engineering design problems (pressure vessel, welded beam).

Conclusions:

• The SCAGTO algorithm effectively addresses AGTO's limitations, offering superior optimization performance.
• SCAGTO exhibits strong engineering practicality and solution advantages for complex optimization tasks.