資源描述:
《基于量子粒子群優(yōu)化算法的壓縮感知數(shù)據(jù)重構(gòu)方法.pdf》由會(huì)員上傳分享,免費(fèi)在線閱讀,更多相關(guān)內(nèi)容在行業(yè)資料-天天文庫(kù)。
1、第28卷第6期傳感技術(shù)學(xué)報(bào)Vo1.28No.62015年6月CHINESEJOURNALOFSENSORSANDACTUATORSJune2015PerceptualDataReconstructionforCompressedSensingBasedonQuantumBehavedParticleSwarmOptimizationL/UZhouzhou,LIYanping(1.Xi’anAeronauticalUniversity,Xi~an710077,China;2.DepartmentofComput
2、erandInformationEngineering,HezeUniversity,HezeShandong274015,China)Abstract:Accordingtowirelesssensornetworkmonitoringobjectfeatures,thecompressedsensingtheoryisappliedtodatacompressiontoreducethecommunicationenergy.Consideringthatreconstructionaccuracyofe
3、xistingdatareconstructionincompressedsensingcanbeeasilyinfluencedbysparsity,afteranalysisofcompressedsensingdatareconstructionprinciple,withsub-flameprocessingtheoriginalsignalinfixedlengthtoreducethesolutionspace,andapplyingquantumtheoryencodinginParticleS
4、warmOptimization,CompressedSensingDataReconstructionthatbasedonQuantum—behavedParticleSwarmOptimizationappears.Accordingtowirelesssensornetworkmonito-ringobjectfeatures,thisalgorithmimprovestheaccuracyofthedatareconstructionbyimprovingparticleinitialpo—siti
5、onandupdatemodeinParticleSwarmOptimizationfromStatistics.Simulationresultsshowthatunderconditionsofsparsitylessthan50,QP—CSDRgets20%~40%performanceimprovementonReconstructionaccuracycomparingtoexistingalgorithms.Nowthealgorithmhasbeenappliedtomicro—earthqua
6、kesandaudiomonitoringsystem,andinactualinspection,theactualsystemlifeisextendedabout2-4timeswithassurancedataaccuracy.Keywords:wirelesssensornetwork;quantumtheory;particleswarmoptimizationalgorithm;compressedsensor;datareconstrueti0nEEACC:7230doi:10.3.issn.
7、1004—1699.2015.06.011基于量子粒子群優(yōu)化算法的壓縮感知數(shù)據(jù)重構(gòu)方法術(shù)劉洲洲,李艷平(1.西安航空學(xué)院,西安710077;2.菏澤學(xué)院計(jì)算機(jī)與信息工程系,山東菏澤274015)摘要:針對(duì)傳感器監(jiān)測(cè)對(duì)象特點(diǎn),將壓縮感知理論應(yīng)用于數(shù)據(jù)壓縮過(guò)程以降低通信能耗,并根據(jù)現(xiàn)有壓縮感知數(shù)據(jù)重構(gòu)算法存在的重構(gòu)精度受稀疏度影響較大的缺點(diǎn),在分析了壓縮感知數(shù)據(jù)重構(gòu)原理后,提出了將原始信號(hào)按固定長(zhǎng)度進(jìn)行分幀處理以減少算法解空間的數(shù)量,并將量子理論中的編碼方式應(yīng)用于粒子群優(yōu)化算法,提出了基于量子粒子群優(yōu)化算法的壓縮
8、感知數(shù)據(jù)重構(gòu)方法QP—CSDR。算法根據(jù)傳感器監(jiān)測(cè)對(duì)象特點(diǎn),從統(tǒng)計(jì)學(xué)角度出發(fā)對(duì)粒子群優(yōu)化算法中的粒子初始位置及粒子群更新方式加以改進(jìn),以提高數(shù)據(jù)重構(gòu)精度。仿真實(shí)驗(yàn)結(jié)果表明,在稀疏度小于5O的條件下,QP—CSDR算法相對(duì)已有算法在重構(gòu)精度方面性能提升20%~40%,該算法已應(yīng)用于微地震及音頻監(jiān)測(cè)系統(tǒng)中,經(jīng)實(shí)際檢驗(yàn)算法在保證數(shù)據(jù)精度的前提下延長(zhǎng)系統(tǒng)壽命2倍~4倍左右。關(guān)鍵詞:量子理論;粒子群優(yōu)化算法;