北大中文核心期刊国外电子测量技术DOI:10.19652/j.cnki.femt.2204414散度值分析的PPG信号伪差滤除算法赖源海孙斌陈运(南京邮电大学自动化学院、人工智能学院南京210023)摘要:光电容积脉搏波(photoplethysmography,PPG)信号的采集极易受到运动伪差的干扰,为了增加信号特征提取的准确度进而提高人体生理参数计算的准确率,提出了基于散度值分析的PPG信号运动伪差滤除算法。将采集的模板信号与实验信号进行带通滤波预处理之后,利用未受伪差干扰的模板信号计算出信号特征的散度值标准阈值范围,接着计算模板信号与受伪差干扰的实验信号的散度值,识别判断并剔除实验信号中存在运动伪差干扰的信号周期,整合得到未受运动伪差干扰的优质波信号。通过人体血管收缩压的检测实验,证明了算法在可穿戴运动系统中的实用性与可靠性。关键词:PPG信号;散度值分析;运动伪差;优质波提取中图分类号:TN98;R318文献标识码:A国家标准学科分类代码:310.61MotionartifactsfilteringmethodforPPGsignalbasedondivergenceanalysisLaiYuanhaiSunBinChenYun(SchoolofAutomationandArtificialIntelligence,NanjingUniversityofPostsandTelecommunications,Nanjing210023,China)Abstract:TheacquisitionofPPGsignalisvulnerabletotheinterferenceofmotionartifact.Inordertoincreasetheaccuracyofsignalfeatureextractionandimprovetheaccuracyofhumanphysiologicalparametercalculation,amotionartifactfilteringalgorithmforPPGsignalbasedondivergencevalueanalysisisproposed.Afterthecollectedtemplatesignalandtheexperimentalsignalarepreprocessedbyband-passfiltering,thestandardthresholdrangeofthedivergencevalueofthesignalcharacteristicsiscalculatedusingthetemplatesignalthatisnotinterferedbythemotionartifacts.Then,thedivergencevalueofthetemplatesignalandtheexperimentalsignalthatisinterferedbythemotionartifactsiscalculated,andthesignalcyclewiththemotionartifactsinterferenceintheexperimentalsignalisidentified,judgedandeliminated,andthehigh-qualitywavesignalthatisnotinterferedbythemotionartifactsisintegrated.Thepracticabilityandreliabilityofthealgorithminwearablesportssystemareprovedthroughthedetectionexperimentofhumanvascularsystolicpressure.Keywords:PPGsignal;divergencevalueanalysis;motionartifact;highqualitywav...