摘要
无人机在完成"最后一公里"的货物配送时需解决任务分配与路径规划问题.本文将"区块链"思想引入拍卖算法中,对无人机编队的任务分配进行优化计算,计算方式由集成中心式计算转变为分布式多智能体间互联计算,使无人机编队在任务计划过程中重新对不合理的任务结果作出调整,使其总回报奖励更高,即分配总成本最少.在确定无人机配送的初末位置后,以路径长度、地形、雷达威胁、无人机碰撞为约束建立目标函数模型,利用改进的量子粒子群算法进行求解.与传统方法相比,本文提出的任务分配策略和路径规划方法可以得到更好的优化结果,并减少计算资源消耗.仿真结果表明:所提出的两种方法在计算效率和任务执行方面都是非常有效性的.
Task assignment and path planning should be solved in"last mile"cargo delivery and pickup for unmanned aerial vehicles(UAVs).This article introduces the idea of"blockchain",an auction algorithm that optimizes the task allocations of drone formation.Blockchain converts integrated centralized computing to distributed multi-agent interconnection computing,which make the UAVs readjust the unreasonable mission results during the mission planning process,and reduces the decision calculation time and costs of drone transportation.After determining the initial and final positions of drone delivery,we establish a cost function model with constraints on path length,terrain,radar threat,and drone collisions resulting in an improved quantum particle swarm algorithm(QPSO).The task allocation strategy and path planning method proposed in this paper can achieve better results than traditional methods,and thereby reduce computing resource consumption.The simulation results from our analysis show that the proposed two methods in terms of computational efficiency and task execution is very effective.
作者
郭兴海
计明军
温都苏
张鑫
田爽
GUO Xinghai;JI Mingjun;WEN Dusu;ZHANG Xin;TIAN Shuang(School of Transportation Engineering,Dalian Maritime University,Dalian 116026,China;Dalian Army Institute of PLA,Dalian 116026,China)
出处
《系统工程理论与实践》
EI
CSSCI
CSCD
北大核心
2021年第4期946-961,共16页
Systems Engineering-Theory & Practice
基金
国家自然科学基金(71971035,71572022)
大连海事大学“双一流”建设专项资金(BSCXXM017)
大连市智慧交通与港行物流重点实验室。
关键词
最后一公里
区块链
任务分配
量子粒子群算法
路径规划
last mile
blockchain
task assignment
quantum particle swarm algorithm
path planning
