第2卷第1期Vol.2No.12023年2月JournalofArmyEngineeringUniversityofPLAFeb.2023差分进化花朵授粉算法的WSN节点部署策略常宇飞1,李重阳1,张爱军1,宋彬杰2(1.32153部队,河北张家口075100;2.陆军炮兵防空兵学院,河南郑州450002)摘要:针对无线传感器网络(wirelesssensornetwork,WSN)的节点部署问题,提出了一种基于差分进化算法(differentialevolutionalgorithm,DEA)优化的花朵授粉算法(flowerpollinationalgorithm,FPA):DE-FPA。设计了动态转换概率,自适应平衡全局授粉和局部授粉间的相互转换,提高算法全局搜索能力。优化了全局授粉过程中的步长缩放因子,进一步提高算法收敛速度。为避免算法陷入局部极值,在每次全局授粉或者局部授粉迭代后引入差分进化策略,增加种群多样性,提高了算法搜索能力。实验结果表明,DE-FPA收敛速度快、寻优精度高,能够在网络连通的约束条件下,达到较高的网络覆盖率。关键词:无线传感器网络;差分进化算法;花朵授粉算法;网络覆盖率中图分类号:TP301.6DOI:10.12018/j.issn.2097-0730.20220703001StrategyofWSNNodeDeploymentBasedonDifferential-EvolutionFlower-PollinationAlgorithmCHANGYufei1,LIChongyang1,ZHANGAijun1,SONGBinjie2(1.Unit32153ofPLA,Zhangjiakou075100,China;2.ArmyArtillery&AirDefenseAcademy,Zhengzhou450002,China)Abstract:Aimingattheproblemofthenodedeploymentofwirelesssensornetwork(WSN),theDE-FPA,anoptimizedflowerpollinationalgorithm(FPA)basedondifferentialevolutionalgorithm(DEA),isproposed.Firstly,theoptimizedalgorithmdesignedthedynamicself-adaptiveconversionprobabilitytobalancethemutualconversionbetweentheglobalpollinationandthelocalpollinationsoastoimprovetheglobalsearchabilityofthealgorithm.Secondly,thestep-sizescalingfactorintheglobalpollinationprocesswasoptimizedtofurtherimprovethealgorithmconvergencespeed.Finally,inordertopreventthealgorithmfromeasilyfallingintothelocalextremevalue,thestrategyofdifferentialevolutionwasintroducedaftereveryglobalpollinationorlocalpollinationiterationtoincreasethediversityofthepopulationandimprovetheal-gorithm'ssearchability.TheexperimentalresultsshowthattheDE-FPAalgorithmhashighernetworkcoverageundertheconstraintsofnetworkconnectivity,withafastconvergencespeedandhighoptimiza...
