ISSN1006-7167CN31-1707/TRESEARCHANDEXPLORATIONINLABORATORY第41卷第12期Vol.41No.122022年12月Dec.2022DOI:10.19927/j.cnki.syyt.2022.12.027基于改进GA-PSO算法的三维WSN气体源定位研究黄浪尘,许诺,张诚(湖南工业大学电气与信息工程学院,湖南株洲412007)摘要:在三维空间针对不同风场下采用无线传感器网络(WSN)定位泄漏气体源问题,提出一种基于风场特征与bounding-box方法的改进遗传-粒子群(GA-PSO)算法。利用bounding-box方法对粒子群搜索空间进行优化处理,减少GA-PSO算法随机搜索产生的无效运算;根据气体湍流扩散模型模拟不同风场特征下气体浓度分布,提出风速修正权重,提高GA-PSO算法搜索目的性。在不同三维风场下将改进GA-PSO算法与多种气体源定位算法进行对比,仿真结果表明,改进GA-PSO算法得到的定位点在无风条件下的平均相对误差为2.66%,有风条件下的平均相对误差为2.84%,定位准确度普遍优于常规算法。关键词:无线传感器网络;改进GA-PSO算法;气体源定位;三维风场特征;气体湍流扩散模型中图分类号:TP391文献标志码:A文章编号:1006-7167(2022)12-0138-06Researchon3DWSNGasSourceLocationBasedonImprovedGA-PSOAlgorithmHUANGLangchen,XUNuo,ZHANGCheng(SchoolofElectricalandInformationEngineering,HunanUniversityofTechnology,Zhuzhou412007,Hunan,China)Abstract:Animprovedgeneticparticleswarmoptimization(GA-PSO)algorithmbasedonwindfieldcharacteristicsandboundingboxmethodisproposedtolocatetheleakagegassourceusingwirelesssensornetwork(WSN)in3-dimensionalspace.Theboundingboxmethodisusedtooptimizetheparticleswarmoptimizationsearchspace,andtoreducetheinvalidoperationcausedbyrandomsearchofGA-PSOalgorithm.Accordingtothegasturbulencediffusionmodel,thegasconcentrationdistributionunderdifferentwindfieldcharacteristicsissimulated,andthewindspeedcorrectionweightisproposedtoimprovethesearchpurposeofGA-PSOalgorithm.TheimprovedGA-PSOalgorithmiscomparedwithavarietyofgassourcelocationalgorithmsunderdifferent3-dimensionalwindfields.ThesimulationresultsshowthattheaveragerelativeerrorofthelocationpointoftheimprovedGA-PSOalgorithmis2.66%underwindlessconditionsand2.84%underwindyconditions.Thelocationaccuracyisgenerallybettert...