A Multi-strategy Improved Osprey Optimization Algorithm for UAV path Planning
Long Yang, Weimin Wang, Binzi Xu · 2024
Addressing the challenges of reduced population diversity and susceptibility to local optima in the later iterations of the Osprey algorithm, this paper presents a Multi-Strategy Improved Osprey Algorithm (IOOA). Firstly, Piecewise mapping is employed to initialize the population, providing a robust foundation for global search. Secondly, during the exploration stage, we design an adaptive parameter Levy exploration method based on the fishhawk leader. Additionally, we integrate a random disturbance strategy inspired by trigonometric functions to enhance population diversity and improve global exploration capacity. During the exploring phase, we propose nonlinear control parameters to effectively balance exploration and exploitation, thus decreasing the risk of the algorithm stagnating to local optimum. Additionally, we introduce an update formula based on the positions of elite leaders to expedite convergence. Finally, we conduct simulation experiments and statistical tests on 13 benchmark test functions, comparing the performance of IOOA with seven popular algorithms. Simultaneously, we apply IOOA to the UAV path planning problem to validate its optimization performance in practical scenarios. The results indicate that IOOA exhibits faster convergence speed and higher effectiveness compared to other optimizers.