Unmanned aerial vehicles (UAVs) performing transportation tasks in complex mountainous areas face challenges from unstructured terrain and the need for dynamic obstacle avoidance. The Grey Wolf Optimization (GWO) algorithm is characterized by its simple structure and minimal parameter tuning, and has demonstrated strong performance in practical applications. However, it suffers from slow convergence speed and a strong tendency to become trapped in local optima. Therefore, this study proposes an Adaptive Intelligent Grey Wolf Optimization (AIGWO) algorithm for UAV transportation path planning in complex mountainous environments. To objectively characterize the mountainous environment and quantitatively assess path quality, a three-dimensional spatial model integrating static terrain, dynamic obstacles, and random noise, together with a multi-objective evaluation function considering path length, flight altitude, and turning angle, were constructed. In order to accelerate convergence, an adaptive search strategy was developed to dynamically balance global exploration and local exploitation. For enhancing the algorithm's exploitation capability and improving population diversity, a candidate position update strategy based on dimensional learning was proposed. Furthermore, dynamic obstacle evolution models and random noise interference mechanisms were established to rigorously evaluate the algorithm's robustness. Finally, the proposed AIGWO algorithm was compared with five advanced algorithms (namely GWO, IGWO, LGWO, PSO, and GA). The results demonstrate that AIGWO achieves a reasonable running time while reducing the number of convergence iterations by 38.8%, shortening the flight path length by 4.6%, and improving the optimal fitness value by 20.0% compared to the benchmark algorithms. These findings confirm the significant superiority of the proposed algorithm for UAV transportation path planning in complex mountainous areas.
Zhou et al. (Sun,) studied this question.