Remaining useful life prediction of lithium-ion battery based on auto-regression and particle filter
基于自回归和粒子滤波的锂离子电池剩余寿命预测
出版日期:2021
杂志名称:International Journal of Intelligent Computing and Cybernetics
卷号:14
期数:2
刊号:1756-378X
出版日期:2021
简介:Withtherapiddevelopmentandstableoperatedapplicationoflithium-ionbatteriesusedinuninterruptiblepowersupply(UPS),thepredictionofremainingusefullife(RUL)forlithium-ionbatteryplayedanimportantrole.Moreandmoreresearcherspaidmoreattentionsonthereliabilityandsafetyforlithium-ionbatteriesbasedonpredictionofRUL.Thepurposeofthispaperistopredictthelifeoflithium-ionbatterybasedonautoregressionandparticlefiltermethod.,Inthispaper,asimpleandeffectiveRULpredictionmethodbasedonthecombinationmethodofauto-regression(AR)time-seriesmodelandparticlefilter(PF)wasproposedforlithium-ionbattery.Theproposedmethoddeformedthedouble-exponentialempiricaldegradationmodelandreducedthenumberofparametersforsuchmodeltoimprovetheefficiencyoftraining.ByusingthePFalgorithmtotracktheprocessoflithium-ionbatterycapacitydeclineandmodifiedobservationsofthestatespaceequations,theproposedPF?+?ARmodelfullyconsideredthedeclinedprocessofbatteriestomeetmoreaccuratepredictionofRUL.,ExperimentsonCALCEdatasethavefullycomparedtheconventionalPFalgorithmandtheAR?+?PFalgorithmbothonoriginalexponentialempiricaldegradationmodelandthedeformeddouble-exponentialone.ExperimentalresultshaveshownthattheproposedPF?+?ARmethodimprovedthepredictionaccuracy,decreasestheerrorrateandreducestheuncertaintyrangesofRUL,whichwasmoresuitableforthedeformeddouble-exponentialempiricaldegradationmodel.,IntherunningofUPSdevicebasedonlithium-ionbattery,theproposedAR?+?PFcombinationalgorithmwillquickly,accuratelyandrobustlypredicttheRULoflithium-ionbatteries,whichhadastrongapplicationvalueinthestableoperationoflaboratoryandotherapplicationscenarios.
