Quadcopter UAV flight trajectory optimization using improved A* algorithm with adaptive heuristic strategy

Abstract

This study proposes a global trajectory planning solution for Quadcopter UAVs to optimize execution time, the number of nodes to be expanded, and ensure mission feasibility and safety. Recognizing the computational performance limitations of traditional algorithms such as Dijkstra, the study proposes an improved method based on the A* algorithm combined with an adaptive heuristic strategy. The core of the method is a flexible conversion mechanism between Manhattan and Euclidean standards based on the actual distance threshold to the target, minimizing the search space and increasing execution speed. To ensure physical feasibility, we integrate a safety boundary mechanism to minimize collision risk and Cubic Spline interpolation techniques to smooth discrete trajectories, ensuring continuity in velocity and acceleration. Simulation scenarios performed on MATLAB R2023b with medium to high obstacle densities have demonstrated the superior effectiveness of the proposed method. Quantitative results show that the proposed algorithm minimizes the number of expansion-required nodes compared to A* and Dijkstra, while significantly improving execution time to meet stringent real-time requirements, creating a solid foundation for deployment on embedded UAV systems in complex 3D environments.