Path planning and engineering problems of 3D UAV based on adaptive coati optimization algorithm.

Area of Science:

• Computational Intelligence
• Metaheuristic Optimization
• Algorithm Design

Background:

• The Coati Optimization Algorithm (COA) faces challenges like imbalanced exploration-exploitation, slow convergence, and local optima entrapment.
• Addressing these limitations is crucial for improving the efficiency and effectiveness of metaheuristic optimization algorithms.

Purpose of the Study:

• To introduce an enhanced variant, the Adaptive Coati Optimization Algorithm (ACOA), designed to overcome the COA's limitations.
• To improve exploration-exploitation balance, convergence speed, and accuracy in optimization problems.

Main Methods:

• ACOA integrates chaos mapping for enhanced randomness and global search.
• A dynamic antagonistic learning approach with random protons mitigates premature convergence.
• An Adaptive Levy Flight strategy maintains population diversity to prevent local optima entrapment.
• A cosine disturbance-based differential evolution strategy eliminates underperforming individuals.

Main Results:

• ACOA demonstrated a superior balance between exploration and exploitation.
• Experimental results on CEC-2017 and CEC-2022 benchmarks showed ACOA outperforming COA and other state-of-the-art algorithms.
• Wilcoxon rank-sum tests confirmed ACOA's statistically significant performance improvements.
• ACOA proved effective on real-world engineering challenges and UAV path planning.

Conclusions:

• The proposed ACOA effectively addresses the limitations of the original COA.
• ACOA offers enhanced robustness, convergence accuracy, and global search capabilities.
• ACOA shows significant potential for complex optimization problems and real-world applications.