王瑞等：基于改进鸽群优化和马尔可夫链的多无人机协同搜索方法 ·1349. 20 (a) -UAVI b UAVI UAV2 UAV3 「A3 UAV4 目标1 目标 10 20 10 15 20 x/km x/km 100 80 40 女PIO 20 -ACO --MSAPIO 30 0 120 150 UAVs运动步数/步 图5无人机的运动轨迹和覆盖率.(a)30步的运动轨迹：(b)60步的运动轨迹：(c)搜索范围覆盖率 Fig.5 The movement trajectories of unmanned aerial vehicle:(a)the movement trajectory of 30 steps:(b)the movement trajectory of 60 steps;(c) search coverage 10 100 (a) b 9 ▣PIO 80 ▣AC0 □MSAPIO 60 6 40 6-PI0 4 30 --ACO ◆-MSAPIO 20 3 10 20 40 6080100120140 160 30 60 90 120 150 无人机运动步数步 搜索时间s 图6协同搜索目标数和有效性评估.(a)搜索目标平均数：(b)有效性评估 Fig.6 Effectiveness evaluation of collaborative search:(a)the average target number;(b)effectiveness assessment 入鸽群优化算法的两个算子中，并使用模拟退火机 (邱华鑫，段海滨。从鸟群群集飞行到无人机自主集群编队 制保留次优个体，避免陷入局部最优 工程科学学报，2017,39(3)：317) [4] Saadaoui H,El Bouanani F.Information sharing based on local 参考文献 PSO for UAVs cooperative search of unmoved targets//2018 In- ternational Conference on Adranced Communication Technologies [1] Qiu H X,Wei C,Dou R,et al.Fully autonomous flying:from and Netcorking (CommNet).Marrakech,Morocco,2018:1 collective motion in bird flocks to unmanned aerial vehicle autono- [5]Yang F,Ji X L,Yang C W,et al.Cooperative search of UAV mous swarms.Sci China Inf Sci,2015,58(12):128201 swarm based on improved ant colony algorithm in uncertain envi- [2] Peng H,Huo M L,Liu ZZ,et al.Simulation analysis of coopera- ronment /2017 IEEE International Conference on Unmanned Sys- tive target search strategies for multiple UAVs /The 27th Chinese tems.Beijing,2017:231 Control and Decision Conference.Qingdao,2015:4855 [6]Duan H B,Qiao PX.Pigeon-inspired optimization:a new swarm [3] Qiu H X,Duan H B.From collective flight in bird flocks to un- intelligence optimizer for air robot path planning.Int J Intell Com- manned aerial vehicle autonomous swarm formation.Chin /Eng, put Cybern,2014,7(1):24 2017,39(3):317 [7]Duan H B.Ye F.Progresses in pigeon-inspired optimization algo-王 瑞等: 基于改进鸽群优化和马尔可夫链的多无人机协同搜索方法 图 5 无人机的运动轨迹和覆盖率. (a)30 步的运动轨迹;(b)60 步的运动轨迹;(c)搜索范围覆盖率 Fig. 5 The movement trajectories of unmanned aerial vehicle: (a) the movement trajectory of 30 steps;(b) the movement trajectory of 60 steps;(c) search coverage 图 6 协同搜索目标数和有效性评估. (a)搜索目标平均数;(b)有效性评估 Fig. 6 Effectiveness evaluation of collaborative search:(a) the average target number; (b) effectiveness assessment 入鸽群优化算法的两个算子中,并使用模拟退火机 制保留次优个体,避免陷入局部最优. 参 考 文 献 [1] Qiu H X, Wei C, Dou R, et al. Fully autonomous flying: from collective motion in bird flocks to unmanned aerial vehicle autono鄄 mous swarms. Sci China Inf Sci, 2015, 58(12): 128201 [2] Peng H, Huo M L, Liu Z Z, et al. Simulation analysis of coopera鄄 tive target search strategies for multiple UAVs / / The 27th Chinese Control and Decision Conference. Qingdao, 2015: 4855 [3] Qiu H X, Duan H B. From collective flight in bird flocks to un鄄 manned aerial vehicle autonomous swarm formation. Chin J Eng, 2017, 39(3): 317 (邱华鑫, 段海滨. 从鸟群群集飞行到无人机自主集群编队. 工程科学学报, 2017, 39(3): 317) [4] Saadaoui H, El Bouanani F. Information sharing based on local PSO for UAVs cooperative search of unmoved targets / / 2018 In鄄 ternational Conference on Advanced Communication Technologies and Networking (CommNet). Marrakech, Morocco, 2018: 1 [5] Yang F, Ji X L, Yang C W, et al. Cooperative search of UAV swarm based on improved ant colony algorithm in uncertain envi鄄 ronment / / 2017 IEEE International Conference on Unmanned Sys鄄 tems. Beijing, 2017: 231 [6] Duan H B, Qiao P X. Pigeon鄄inspired optimization: a new swarm intelligence optimizer for air robot path planning. Int J Intell Com鄄 put Cybern, 2014, 7(1): 24 [7] Duan H B, Ye F. Progresses in pigeon鄄inspired optimization algo鄄 ·1349·