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

In response to the challenges faced by the Coati Optimization Algorithm (COA), including imbalance between exploration and exploitation, slow convergence speed, susceptibility to local optima, and low convergence accuracy, this paper introduces an enhanced variant termed the Adaptive Coati Optimization Algorithm (ACOA). ACOA achieves a balanced exploration-exploitation trade-off through refined exploration strategies and developmental methodologies. It integrates chaos mapping to enhance randomness and global search capabilities and incorporates a dynamic antagonistic learning approach employing random protons to mitigate premature convergence, thereby enhancing algorithmic robustness. Additionally, to prevent entrapment in local optima, ACOA introduces an Adaptive Levy Flight strategy to maintain population diversity, thereby improving convergence accuracy. Furthermore, underperforming individuals are eliminated using a cosine disturbance-based differential evolution strategy to enhance the overall quality of the population. The efficacy of ACOA is assessed across four dimensions: population diversity, exploration-exploitation balance, convergence characteristics, and diverse strategy variations. Ablation experiments further validate the effectiveness of individual strategy modules. Experimental results on CEC-2017 and CEC-2022 benchmarks, along with Wilcoxon rank-sum tests, demonstrate superior performance of ACOA compared to COA and other state-of-the-art optimization algorithms. Finally, ACOA's applicability and superiority are reaffirmed through experimentation on five real-world engineering challenges and a complex urban three-dimensional unmanned aerial vehicle (UAV) path planning problem.