Application of the differential evolution algorithm in unmanned aerial vehicle (UAV) pathfinding

Abstract

This paper studies the differential evolution (DE) algorithm for UAV path planning in a three-dimensional complex environment. By simulating the mutation, crossover and selection operations in the natural evolution process, the flight path of the UAV is gradually optimized. The experiment compares the performance of the DE algorithm with other path planning algorithms under the same starting point, end point and obstacle settings, focusing on the path length and calculation time. The experiment shows that the path optimization quality of the DE algorithm is significantly improved, and it can effectively avoid obstacles, which is suitable for application scenarios that require high-precision path planning. The SHADE algorithm is also used to experiment with the multi-path patrol conflict problem.