文档名：最小二乘算法优化及其在锂离子电池参数辨识中的应用
摘要：传统最小二乘法(LS)用于锂离子电池模型在线参数辨识精度低,通过带遗忘因子递推最小二乘算法能够有效地提高辨识精度,但固定的遗忘因子影响模型动态特性.遗忘因子的自适应处理能提高算法对动态系统的参数辨识能力,而目前的自适应方法容易忽略模型参数的稳定性,同时方法待定系数范围较大且难以确认.为了得到高精度且稳定性良好的模型参数,该文设计了一种精度和稳定性兼优且更简单的自适应遗忘因子递推最小二乘(AFFRLS)改进方法,并与其他AFFRLS、可变遗忘因子递推最小二乘(VFFRLS)进行仿真对比分析.结果表明,改进的AFFRLS能够在模型精度和参数稳定性取得更好的平衡,且对不同的在线工况具有良好的适用性.
Abstract：Offlineandonlinemethodsareusedtoidentifymodelparameters,butthemodeldynamiccharacteristicobtainedbytheonlinemethodisbetter.Therecursiveleastsquaresmethodissimpleandoftenusedforonlineparameteridentificationoflithium-ionbatterymodels.However,theleastsquaremethod(RLS)hasalowidentificationaccuracy.Thus,theforgettingfactorrecursiveleastsquaremethodwasproposedtoimprovetheaccuracyofparameteridentification.Toimprovethedynamicidentificationability,thevariableforgettingfactorleastsquare(VFFRLS)methodandadaptiveforgettingfactorrecursiveleastsquare(AFFRLS)methodappear.Yetthecurrentadaptivemethodstendtoignorethestabilityofmodelparameters,andtheundeterminedcoefficientrangeofthismethodislargeanddifficulttoconfirm.Themodelparameterchangesdrastically,anditiseasytocausethedivergenceofthealgorithm.ThispaperproposesasimplerAFFRLSmethodwithoutanundeterminedcoefficienttoaddresstheseissues.Andittakesintoaccounttheaccuracyandstabilityofthemodel.Firstly,basedondynamicstresstesting(DST)andFederalCityOperatingConditions(FUDS)data,theFFRLSmethodwithfixedforgettingfactorvalueissimulatedandanalyzed,andtheinfluencetrendofdifferentforgettingfactorsontheaccuracyandstabilityofmodelparametersisobtained.Secondly,theproposedAFFRLSmethodiscomparedwithotherAFFRLSandVFFRLS,andthestabilityandaccuracyoftheidentificationparametersareanalyzed.Finally,theerrortrackingabilityandconvergencespeedofthethreeadaptivemethodsareanalyzed,andtheadaptiveperformanceoftheproposedAFFRLStoDSTandFUDSconditionsareanalyzed.TheFFRLSsimulationresultswithfixedforgettingfactor(λ)valueshowthatwhenλvaluedecreases,thealgorithmhasbettertrackingabilityfortime-varyingparameters,theconvergencespeedisaccelerated,andtheidentificationaccuracyiseffectivelyimproved.However,whentheλvaluedecreases,theparameterchangesdrastically,andthestabilitydecreases.Itcanbeseenthatobtainingtheappropriateλvalueisimportantfortheidentificationabilityoftheadaptivemethods.TheresultsofthethreeadaptivemethodssimulationsshowthattheimprovedAFFRLSinthispaperhasbettertrackingabilityfortime-varyingparametersandhighmodelaccuracy.AndithasbetterstabilityoftheparameterobtainedbyFFRLSwithfixedλvaluesof0.980and0.985.ItcanbeseenthattheproposedAFFRLScanachieveabetterbalancebetweenaccuracyandstability.TherelationshipbetweentheλvalueandtheerroroftheadaptivemethodsshowsthattheimprovedAFFRLScantracktheerrorvariationbetter.Bycomparingtheoperationtimewiththethreemethods,theresultsshowthattheproposedAFFRLShasafasterconvergencerate.AccordingtotherelationshipbetweenλvalueandtimeinDSTandFUDSconditions,theimprovedAFFRLSmethodhasthemajorityofλvaluenear0.980intheFUDScondition,andthemajorityofλvalueis1intheDSTcondition.Thesimulationanalysisshowsthat:(1)TheproposedAFFRLSmethodcanimprovetheaccuracyofthealgorithmandtakethestabilityofmodelparametersintoconsideration,andithasagoodbalancebetweenalgorithmaccuracyandparameterstability.ApplyingtheproposedAFFRLSmethodandKalmanfiltertopredictthestateofchargecanimprovethepredictionaccuracy.(2)TheproposedAFFRLSmethodhasbettertrackingabilityforerrorvariationandfasterconvergencespeed.(3)Theproposedmethodcanimprovethealgorithm'saccuracyunderbothslowanddrasticconditions,soit'ssuitablefordifferentonlineconditions.
作者：范兴明 封浩 张鑫Author：FanXingming FengHao ZhangXin
作者单位：桂林电子科技大学机电工程学院桂林541004
刊名：电工技术学报
Journal：TransactionsofChinaElectrotechnicalSociety
年，卷(期)：2024, 39(5)
分类号：TM912
关键词：锂离子电池模型 参数辨识 最小二乘法 自适应遗忘因子
Keywords：Lithium-ionbatterymodel parameteridentification leastsquaremethod adaptiveforgettingfactor
机标分类号：TP242TM914TP391.41
在线出版日期：2024年3月19日
基金项目：国家自然科学基金，广西自然科学基金项目最小二乘算法优化及其在锂离子电池参数辨识中的应用[
期刊论文] 电工技术学报--2024, 39(5)范兴明 封浩 张鑫传统最小二乘法(LS)用于锂离子电池模型在线参数辨识精度低,通过带遗忘因子递推最小二乘算法能够有效地提高辨识精度,但固定的遗忘因子影响模型动态特性.遗忘因子的自适应处理能提高算法对动态系统的参数辨识能力,而目前...参考文献和引证文献
参考文献
引证文献
本文读者也读过
相似文献
相关博文
最小二乘算法优化及其在锂离子电池参数辨识中的应用 Optimization of Least Squares Method and Its Application in Parameter Identification of Lithium-Ion Battery Model
最小二乘算法优化及其在锂离子电池参数辨识中的应用.pdf
- 文件大小:
- 13.73 MB
- 下载次数:
- 60
-
高速下载
|
|
