A Hybrid Large Neighborhood Search Algorithm for Solving the Multi Depot UAV Swarm Routing Problem

This paper focuses on a modified Multi-Depot Unmanned Aerial Vehicle Routing Problem (MMDUAVRP). Comparing to classic multi-depot vehicle routing problem, our studied problem has no constraints restrict the depot where (UAV) departs and returns. work aims minimize number of UAVs total distance traveled by all UAVs. is mathematically formulated in this heuristic-assignment based hybrid large neighborhood search(HLNS) proposed solve it. Extensive computational experiments are conducted verify performance HLNS. The HLNS algorithm was first tested Multi Traveling Salesman which simplified version MMDUAVRP, high quality solutions have been obtained. Experimental results compared with CPLEX other well-known algorithms suggest that provides better within comparatively shorter period time. In addition, we also conduct sensitivity analysis location depots task points may affect cost solution.