基于三支决策的无人机-车协同规划算法研究

Research on UAV-Car collaborative planning algorithm based on three-way decisions

随着物流车、无人机技术的逐渐成熟以及在运输中的出众表现,车辆与无人机协同作业的路径规划问题(VRPUAV)成为当前学术和工程界亟待解决的问题.本文基于车辆与无人机协同作业场景,将运输问题划分为三粒度,即"无人机数量为0","无人机数量不足","无人机数量充足".采用蚁群算法和无人机物流车协同运输优化算法对3个子问题分别提出相应的解决策略;最后通过仿真实验证明了算法在行驶成本与时间成本上的优化作用,同时在运输物品总条件一定的前提下,三类子问题的治略方案均能够正确求解出优化解,且第2、3种场景较第1种,第3种场景较第2种具有更优运输成本和客户等待时间成本,充分证明了三粒度分治的合理和有效性.

With the gradual maturity of unmanned vehicles and unmanned aerial vehicle technology and outstanding performance in transportation,the route planning problem(VRPUAV)of vehicles and drones cooperation has become a problem that needs to be solved in the current academic and engineering fields.Based on the collaborative operation scenario of vehicles and drones,the paper divides the transportation problem into three granularities,namely,“the number of drones is 0”,“the number of drones is sufficient,”and“the number of drones is insufficient”,and uses ant colony algorithm and the cooperative optimization algorithm of drone and unmanned vehicle to propose corresponding solutions to each of the three sub-problems.Finally,simulation experiments are conducted to verify the optimization effect of the algorithm on driving cost and time cost.Under a certain premise,the governance solutions for the three sub-problems can correctly solve the optimal solution,and the second and third scenarios have better transportation costs and customer waiting time than the first and third scenarios is also better than the second in the similar variables.The cost has fully proved the reasonableness and effectiveness of the three-way decisions and treatment.