MMPA: A modified marine predator algorithm for 3D UAV path planning in complex environments with multiple threats

In this paper, we propose a modified Marine Predators Algorithm (MPA) for global optimization in complex environments with multiple threats, specifically targeting unmanned aerial vehicle (UAV) path planning. Addressing the shortcomings of the original MPA, we introduce four innovative strategies, including adaptive parameter control, nonlinear inertia weight, Cauchy mutation operator-based randomization, and improved differential mutation strategy. These strategies not only significantly enhance the convergence speed while ensuring algorithm precision but also provide effective avenues for enhancing the performance of MPA. We successfully apply these modifications to UAV path planning scenarios in complex environments. To validate the proposed algorithm, we conduct comprehensive tests using 23 classical benchmark functions and compare its performance with six well-known algorithms. The experimental results demonstrate that MMPA excels in numerical optimization problems with various modes, exhibiting superior optimization performance. Moreover, in eight 3D Unmanned Aerial Vehicle (UAV) 22 path planning scenarios with diverse complexities, we demonstrate the superiority and robustness of MMPA in tackling practical problems. By employing the four innovative strategies, MMPA achieves notable performance improvements in complex tasks, showcasing strong potential for practical applications. Overall, our research not only presents an effective approach to enhance the MPA algorithm's performance but also demonstrates significant advantages in addressing practical problems. These innovative strategies offer valuable insights for advancing the research and application of nature-inspired optimization algorithms.