An Empirical Study on Genetic Algorithm for 2D Path Planning of Unmanned Aerial Vehicles with Obstacle Avoidance
Ya-Hong Zhao, Yulin Lan, Shang-Lin Li, Wenfen Zhang · 2023
Unmanned aerial vehicle (UAV), with high efficiency, low manpower costs, and flexible convenience, has attracted widespread attention in logistics. UAV trajectory planning is an essential component in this context. Genetic algorithm(GA) is an effective tool for solving complex route optimization problems. However, selecting the optimal combination of parameters for UAV path planning remains a challenging problem, especially in complex urban environments with numerous obstacles. This paper proposes a genetic algorithm-based UAV path planning approach(UAV_GA) that considers obstacle avoidance on a two-dimensional grid-based map. Furthermore, to investigate the general patterns in selecting parameters for GA in such problems, a comprehensive empirical study is conducted. The study analyzes the effects of population size, number of iterations, crossover, and mutation probabilities on algorithm performance. It aims to identify universal patterns for optimal parameter combinations and provide guidance for parameter selection in GA for UAV trajectory planning with obstacle avoidance.