Optimization approach to multidrone path planning

Abstract

This paper proposes a novel multi-objective optimization (MOO) approach to cooperative path-planning of multiple unmanned aerial vehicles (UAVs). The cooperative multi-UAV path-planning problem is formulated in a MOO framework to determine optimal trajectories from the given source to the destination, while avoiding collision with obstacles, teammates, and mountains. To address this challenge, a Firefly Algorithm with Non-dominated Sorting (FANS) is developed, offering a robust solution to the MOO problem by facilitating rapid convergence of solutions to the optimal Pareto Front. This makes it particularly suitable for tackling real-world MOO problems. Comprehensive experiments undertaken with diverse workspaces reveal the superiority of the proposed algorithm over its state-of-the-art contenders in the present context. These experiments demonstrate not only the efficiency of the FANS algorithm in achieving optimal path-planning but also its adaptability to varying environmental complexities, highlighting its potential for practical applications in UAV navigation.