[1]邹丽媛1,2,王 宏1,等.基于脑功能网络深度学习的车内噪声评价模型[J].机械与电子,2020,(05):76-80.
 ,,et al.Research on Vehicle Interior Noise Evaluation Model Based on Stacked Antoencoder and Functional Brain Network[J].Machinery & Electronics,2020,(05):76-80.
点击复制

基于脑功能网络深度学习的车内噪声评价模型()
分享到:

机械与电子[ISSN:1001-2257/CN:52-1052/TH]

卷:
期数:
2020年05期
页码:
76-80
栏目:
智能工程
出版日期:
2020-05-22

文章信息/Info

Title:
Research on Vehicle Interior Noise Evaluation Model Based on Stacked Antoencoder and Functional Brain Network
文章编号:
1001- 2257(2020)05- 0076- 05
作者:
邹丽媛王 宏宋桂秋
1.东北大学机械工程与自动化学院,辽宁 沈阳 110819;
2.辽宁水利职业学院,辽宁 沈阳 110122
Author(s):
ZOULiyuanWANGHongSONGGuiqiu
1.SchoolofMechanicalEngineeringandAutomation,NortheasternUniversity,Shenyang110819,China;
2.LiaoningWaterConservancyVocationalCollege,Shenyang110122,China
关键词:
车内噪声评价堆栈自编码器脑功能网络分析同步似然
Keywords:
vehicle interior noise evaluationstack AutoEncoderbrain network analysis synchronouslikelihood
分类号:
R318
文献标志码:
A
摘要:
研究并构建了一个结合脑电信号处理与深度学习的车内噪声评价模型,该算法通过自我学习实现脑电信号特征提取,使用同步似然方法构建delta、alpha和beta频段的脑功能网络。将3个频带的
脑功能网络扁平化处理后作为输入,通过无监督的堆栈自编码器(RSAE)自主提取脑功能网络的特征。通过几个高阶 特 征 训 练 前 后 对 比,证 实 了 RSAE 自 主 学 习 到 与 噪 声 评 价 有 关 的 脑 神 经 特 征。最 终 将 RSAE与普遍使用的 SVM 回归模型进行比较,同时将脑功能网络与传统的基于心理声学声音品质的车内噪声评价进行对比。结果表现,所提出的脑功能网络 RSAE 模型的平均决定系数高达98.69%,明显优于其他方法。
Abstract:
Inthispaper,anin vehiclenoiseevaluation modelcombiningEEGsignalprocessinganddeep learningwasstudiedandconstructed.ThisalgorithmcanrealizeEEGsignalfeatureextractionthroughself learning.Thesynchronouslikelihoodmethodwasusedtoconstructthebrainfunctionnetworkofdelta,alphaandbeta bands.FortheunsteadycharacteristicsofEEGsignals,thesynchronouslikelihoodmethoddoesbetterinfindingthe linearandnonlinearcouplingrelationshipbetweendifferentchannels.Thethreefrequencybandsoffunctionalbrain networkswereflattenedasaninput.Firstly,thefeaturesofthefunctionalbrainnetworkwereextractedthrough unsupervisedpre training,andthentheinitializationweightswereusedtotrainthefour layernetworkusingthe in vehiclenoisescoresevaluatedbytheexpertgroupastheannotationinformation.Throughthecomparisonof severalhigh orderfeaturesbeforeandaftertraining,thevalidityofRSAE’sself learnedfeatureswereconfirmed tobeuseful.Intheend,thispapercomparedtheRSAEwithSVMregressionmodel,andcomparedthefunctional brainnetworkwiththetraditionalpsychoacousticsoundquality basedinteriornoiseevaluation.Theresultsshow thattheproposedvehicleinteriornoisebasedonRSAEandfunctionalbrainnetworkwasbetter.TheaveragedecisioncoefficientofthefunctionalbrainnetworkRSAEmodelproposedisashighas98.69%,whichisobviouslysuperiortoothermethods.Theresultsalsoconfirmedthefeasibilityandeffectivenessoftheproposedmodel.

参考文献/References:

[1] LIZG,DIG Q,JIAL.RelationshipbetweenElectroencephalogramvariationandsubjectiveannoyanceundernoiseexposure[J].Applied Acoustics,2013,75(1):37- 42.

[2] PAVLENKOVB,CHERNYISV,GOUBKINADG.EEGcorrelatesofanxietyandemotionalstabilityinadulthealthysubjects[J].Neurophysiology,2009,41(5):337- 345.
[3] SHIN D,SHIN D,SHIN D.Developmentofemotion recognitioninterfaceusingcomplex EEG/ECG biosignalforinteractivecontents[J].Multimedia Tools andApplications,2016,76(9):1- 22.
[4] SCHAPKINSA,FALKENSTEIN M,MARKSA,et al.Executivebrainfunctionsafterexposuretonocturnaltrafficnoise:effectsoftask difficultyandsleep quality[J].EuropeanJournalofAppliedPhysiology,2006,96(6):693- 702.
[5] 张朕,焦学军,杨涵钧,等.航天噪声环境对脑机接口的影响研究[J].载人航天,2017,23(2):274- 278.
[6] 戚作秋,王宏,常文文,等.生产性噪声对工作人员脑认知影响的 ERP分析[J].东北大学学报(自然科学版),2017,38(11):1590- 1594.
[7] RUDZIKF,THIESSEL,PIERENR,etal.Sleepspindlecharacteristicsandarousabilityfrom night time
transportationnoiseexposureinhealthyyoungandolderindividuals[J].Sleep,2018,41(7):1- 14.
[8] DANG VuTT,MCKINNEYS M,BUXTON O M,etal.Spontaneousbrainrhythmspredictsleepstability inthefaceofnoise[J].CurrentBiology,2010,20(15):626- 627.
[9] TASSIP,ROHMER O,BONNEFOND A,etal.Long term exposureto nocturnalrailway noise produces chronicsignsofcognitivedeficitsanddiurnalsleepiness[J].JournalofEnvironmentalPsychology,2013,33
(33):45- 52.
[10] 李博,李树森,李滨.割灌机作业噪声及其对园林工人脑波的影响[J].北京林业大学学报,2017,39(4):108- 114.
[11] LUCF,TENGS,HUNGCI,etal.Reorganizationof functionalconnectivityduringthe motortaskusing
EEGtime frequencycrossmutualinformationanalysis[J].ClinicalNeurophysiology OfficialJournalof
theInternationalFederationofClinicalNeurophysiology,2011,122(8):1569- 1579.
[12] GARCIAPRIETOJ,BAJO R,PEREDA E.Efficient computationoffunctionalbrainnetworks:towardreal
timefunctionalconnectivity[J].Frontiersin Neuroinformatics,2017,11(8):1- 18.
[13] SAKKALISV.Reviewofadvancedtechniquesforthe estimationofbrainconnectivitymeasuredwithEEG/ MEG[J].ComputersinBiologyand Medicine,2011,41(12):1110- 1117.
[14] VIALATTE F,OISB,MARTIN C,etal.A machine learningapproachtotheanalysisofTime Frequency Maps,andItsApplicationtoNeuralDynamics[J].Neural NetworkstheOfficialJournaloftheInternationalNeural NetworkSociety,2007,20(2):194- 209.
[15] MARIEM G,AMMAR L,RIDHA E,etal.Stacked sparseautoencoderandhistoryofbinarymotionimageforhumanactivityrecognition[J].Multimedia ToolsandApplications,2019,78:2157- 2179.

备注/Memo

备注/Memo:
收稿日期:2019- 12- 23
基金项目:辽宁省高等学校创新团队项目(LT2014006);国家重点研发计划(2017YFB1300300)
作者简介:邹丽媛 (1982-),女,内蒙古赤峰人,博士研究生,研究方向为信号处理。
更新日期/Last Update: 2020-05-22