Isistimu yokuphatha ibhethri ithola isimo sokushajwa nesimo sempilo kumabhethri e-lithium-ion ngokusebenzisa amasu okulinganisa angaqondile. Ayikwazi ukukala lezi zinto ngokuqondile ngoba ibhethri inokusabela okuyinkimbinkimbi ngaphakathi. Ngakho-ke, isistimu isebenzisa izindlela ezifana nokukhipha izici zezibalo, ukubala kwe-Coulomb, namamodeli athuthukisiwe aqhutshwa idatha. Isibonelo, ibheka amamethrikhi ezibalo njengokuhluka, incazelo, kanye nokutsheka kugesi kanye namajika amanje ukuze kubhekwe ukonakala kwebhethri. Ukusebenzisa izindlela zokulinganisa ezingaqondile, njengokufunda komshini nezindlela ezisuselwe kuzibukeli, kusiza ukwenza ukulinganisa kwezenhlalo kunembe kakhudlwana futhi kuphephe. Lezi zindlela zokulinganisa i-soc zisiza isistimu yokuphatha ibhethri ibikezele izinguquko kumabhethri e-lithium-ion. Ziphinde zisize ukuphatha ukulahleka kwamandla, ukuguga, kanye nezingozi kangcono. Isilinganiso esihle sokushaja nesimo sokulinganisela kwezempilo kusiza lonke ibhethri le-lithium-ion lisebenze kangcono futhi lihlale isikhathi eside.
Ukulinganisa okunembile kwe-soc kumasistimu ebhethri ye-lithium-ion kugcina ibhethri liphephile ekushajweni ngokweqile, ekushiseni ngokweqile, nasekuhlulekeni okungazelelwe. Lokhu kwenza amasu okulinganisa aqinile abaluleke kakhulu ezinhlelweni zesimanje zokuphatha ibhethri.
I-Statistical Metric | Incazelo | Ukuhlobana Nokuwohloka Kwebhethri |
|---|---|---|
Ukuhluka | Ihlola ukuthi zimi kanjani izinguquko ze-voltage/zamanje | Ukwehluka okuphezulu kusho ukumelana nokungalingani kwangaphakathi nokusabela kwamakhemikhali, nokulimala kwama-electrode |
Inani Eliphezulu | I-voltage ephezulu kakhulu/yamanje ngesikhathi sokushaja noma ukushajwa | Izinombolo eziphansi zibonisa umthamo omncane wokulayisha kanye nezinkinga zokuphepha ezingaba khona njengokushajwa ngokweqile noma ukushisa ngokweqile |
Inani Elincane | I-voltage ephansi/yamanje ngesikhathi sokushaja noma ukushajwa | Ibonisa ukulahlekelwa amandla kanye nezinkinga zokuphepha |
Isho (Isilinganiso) | I-voltage emaphakathi/yamanje phakathi nomjikelezo | Izinguquko zibonisa ukuwohloka kwe-electrolyte kanye nokukhipha amandla amancane |
Skewness | Yeka ukuthi i-voltage/yamanje isabalala kanjani | Isetshenziswa ekukhishweni kwesici ukubikezela i-SOH |
I-Kurtosis Eningi | Ibukhali kangakanani i-voltage/ukuphakama kwamanje | Izinombolo eziphakeme zisho i-polarization eyengeziwe kanye nekhono elincane lokufaka i-lithium |
Izitoreji Eziyinhloko
Amasistimu okuphatha ibhethri awakwazi ukukala ishaji noma impilo ngokuqondile. Basebenzisa izindlela ezingaqondile njengokuhlaziya izibalo, ukubala kwe-Coulomb, nokufunda komshini. Lezi zindlela zisiza ukulinganisa ukushajwa kwebhethri nempilo.
Ukwazi isimo sokushajwa kusiza ukugcina amabhethri ephephile. Iyeka ukushaja ngokweqile, ukushisa ngokweqile kanye nezinkinga ezizumayo.
Kunezindlela ezahlukene zokuhlola amabhethri. I-Open Circuit Voltage, i-Coulomb Counting, i-Kalman Filtering, namamodeli asekelwe ku-AI ezinye zezindlela. Ngamunye unamaphuzu amahle namabi. Ukuzisebenzisa ndawonye kwenza imiphumela ibe ngcono futhi ithembeke kakhulu.
Isilinganiso sesimo sezempilo sihlola ukuthi ibhethri lineminyaka emingaki. Kubheka ukulahlekelwa amandla kanye nokumelana kwangaphakathi. Lokhu kusiza ukuqagela impilo yebhethri futhi kugweme izinkinga zokuphepha.
Izindlela ezixubile ezihlanganisa izindlela ezisuselwe kumodeli nezindlela eziqhutshwa idatha. Lokhu kunikeza imiphumela engcono kakhulu. Bangashintsha ngokusetshenziswa komhlaba wangempela. Lokhu kusiza amabhethri ukuthi ahlale isikhathi eside futhi asebenze kangcono.
Izisekelo Zesistimu Yokuphatha Ibhethri
Imisebenzi ebalulekile
Uhlelo lokuphatha ibhethri lubaluleke kakhulu kumabhethri e-lithium-ion. Isiza ukugcina amabhethri e-lithium-ion ephephile futhi esebenza kahle. Uhlelo luhlola iseli ngalinye lebhethri le-lithium-ion ukuze lithole i-voltage, yamanje, nezinga lokushisa. Iphinde iqinisekise ukuthi wonke amaseli ebhethri e-lithium-ion ayashaja futhi aphume ngokulinganayo. Lokhu kusiza ibhethri ngalinye le-lithium-ion ukuthi lihlale isikhathi eside futhi lisebenze kangcono.
Isistimu yokuphatha ibhethri ibheka isimo sokushajwa kanye nesimo sempilo ngebhethri ye-lithium-ion ngayinye. Isebenzisa lezi zinombolo ukumisa ukushaja ngokweqile kanye nokukhipha amanzi ajulile, okungalimaza amabhethri e-lithium-ion.
Ukuphepha kuza kuqala. Isistimu izonqamula ibhethri ye-lithium-ion uma ithola izinkinga ezifana nokushisa ngokweqile noma amasekhethi amafushane. Ingasebenzisa amaseli ekhophi yasenqolobaneni noma amaphekhi ukuze igcine izinto zisebenza.
Izindaba zokuxhumana. Isistimu yokuphatha ibhethri isebenzisa ibhasi le-SPI ne-CAN ukuthumela idatha kwezinye izingxenye zedivayisi noma imoto.
Kunezinhlobo ezihlukene, ezifana nendawo eyodwa noma esabalalisiwe, ngakho-ke uhlelo lokuphatha ibhethri lungalingana nemiklamo eminingi yebhethri ye-lithium-ion.
Amanye amasistimu anezici ezengeziwe ezifana nokuqapha okukude, ukuqagela umjikelezo wempilo, nokutholwa kwamaphutha. Lokhu kusebenzisa i-cloud computing nokufunda komshini ukusiza ibhethri lisebenze kangcono futhi liphephe.
Umsebenzi Obalulekile / I-algorithm | Incazelo |
|---|---|
Ukuqapha Iseli | Ibuka i-voltage, yamanje, kanye nezinga lokushisa kweseli yebhethri ye-lithium-ion ngayinye. Ithola izinkinga bese iqala izenzo zokuphepha. Ibonisa isimo senkokhelo kanye nesimo sempilo. |
Ukuthuthukisa Amandla | Ilawula ukushaja nokukhipha ukuze kugcinwe amaseli ebhethri e-lithium-ion ephephile. Isebenza namanye amasistimu ukusebenzisa amandla ngendlela ehlakaniphile. |
Isiqiniseko Sokuphepha | Imisa izingozi njengokubaleka okushisayo. Isebenzisa izinhlelo eziyisipele futhi igcina abantu bephephile ekushayweni ugesi. |
Ukuthuthukisa Ukushaja Ibhethri | Ishintsha ukushaja ukuze kwehlise ingcindezi kuseli yebhethri ye-lithium-ion ngayinye. Ilondoloza amakhodi amaphutha ukuze kuhlolwe kamuva. |
I-Algorithm Yokulinganisa Iseli | Iqinisekisa ukuthi wonke amaseli ebhethri e-lithium-ion ane-voltage efanayo. Isebenzisa ukulinganisa okusebenzayo noma okwenziwayo ukusiza ibhethri lisebenze kangcono. |
Ama-algorithms wokuxhumana | Ithumela idatha phakathi kwesistimu yokuphatha ibhethri namanye amadivaysi. Iyayeka ukushaja uma ithola izimo ezingaphephile. |
Ithiphu: Ukusebenzisa isofthiwe esenziwe kakade namathuluzi wehadiwe kungasiza onjiniyela bakhe futhi bahlole isistimu yokuphatha ibhethri yamabhethri e-lithium-ion ngokushesha.
Amakhemikhali Asekelwe
Isistimu yokuphatha ibhethri idinga ukusebenza namakhemistri webhethri ye-lithium-ion amaningi. Ikhemistri ngayinye, njenge-NMC, LFP, ne-NCA, inamaphuzu ayo amahle namabi. Isibonelo, amabhethri e-lithium-ion e-NMC anomthamo omkhulu wamandla. Amabhethri e-LFP e-lithium-ion ahlala isikhathi eside futhi aphatha ukushisa kangcono. Isistimu yokuphatha ibhethri ishintsha indlela esebenza ngayo ukuze ilingane nekhemistri yebhethri ye-lithium-ion ngayinye.
Ucwaningo lwakamuva lubheka ukuthi amakhemikhali ebhethri e-lithium-ion asebenza kanjani ezimotweni zikagesi. Lezi zifundo zibonisa ukuthi amasistimu okuphatha amabhethri kufanele aphathe izinguquko ekuminyana kwamandla, izindleko, nempilo yomjikelezo. Ziphinde zibonise ukuthi ukuphathwa okushisayo kanye nokulinganisa kombuso okuthuthukisiwe kubalulekile ohlotsheni ngalunye lwebhethri ye-lithium-ion. Amamodeli okufunda omshini angasiza ukubikezela isimo sempilo yamabhethri e-lithium-ion ngokusebenzisa idatha ehlungiwe. Lokhu kwehlisa amaphutha futhi kusiza isistimu yokuphatha ibhethri ibhekane neminyaka yobudala bekhemistri yebhethri ye-lithium-ion.
Uhlelo lokuphatha ibhethri oluguquguqukayo lungasebenza namakhemistri amaningi ebhethri ye-lithium-ion. Lokhu kusiza lonke uhlelo lokusebenza, kusukela ezimotweni zikagesi kuye kweziphathwayo zikagesi, zithole ukusebenza kwebhethri okungcono kakhulu nokuphepha.
Isimo Sokushaja Kumabhethri e-Lithium-Ion
Isimo sokushaja sibaluleke kakhulu kumabhethri e-lithium-ion. Isiza ukugcina ibhethri liphephile futhi lisebenza kahle. Uma isimo sokushajwa singalungile, ibhethri lingashisa kakhulu noma liphelelwe amandla. Lokhu kungenza ibhethri liphuke noma kudale izinkinga eziyingozi njengomlilo. Ezimotweni zikagesi, ukwazi isimo sokushaja kuyasiza ngokubhuleka nokushaja. Futhi yenza ibhethri lihlale isikhathi eside. Ucwaningo lukhombisa ukuthi ukulinganisa okuhle kwezindleko kwehlisa amaphutha futhi kusize imvelo.
Awukwazi ukukala isimo sokushaja ngokuqondile ebhethrini le-lithium-ion. Ukusabela kwamakhemikhali ngaphakathi kufihliwe futhi kunzima ukukubona. Izinzwa zingaba iphutha ngenxa yomsindo nezinguquko ebhethrini. Ngakho, amasistimu okuphatha ibhethri asebenzisa izindlela ezikhethekile ukuqagela isimo sokushaja. Babheka i-voltage, i-current, kanye nezinga lokushisa ukuze bakuqonde. Lezi zindlela zisiza ukubhekana nezinkinga zenzwa nokuguga kwebhethri.
Indlela ye-OCV
Indlela ye-Open Circuit Voltage iqagela isimo sokushaja ngokubheka i-voltage yebhethri ngemva kokuphumula. Ikhemikhali yebhethri ngayinye ine-voltage yayo kanye nesixhumanisi sesimo sokushaja. Le ndlela ilula futhi ayibizi kakhulu. Isebenza kahle uma uqala ukushaja futhi ayidingi imodeli yebhethri enkulu.
aspect | Yekuchumana |
|---|---|
isimiso | I-voltage yebhethri ikalwa ngemuva kokuphumula. I-OCV kanye nesixhumanisi sesimo sokushaja sitholakala ngokuhlola uhlobo lwebhethri ngalunye. |
Izinzuzo | 1. Inqubo elula |
Ukulinganiselwa | 1. Idinga isikhathi eside sokuphumula (ngaphezu kwamahora ama-2 uma kubanda) |
Indlela ye-OCV ayikwazi ukuhlola isimo sokushajwa ngenkathi ibhethri isebenza. Amabhethri e-lithium-ion avame ukushintsha ngokushesha, ngakho ukulinda ukuthi ibhethri liphumule akusizi. Amabala ayisicaba ejikeni le-OCV akwenza kube lula ukuthola amaphutha amakhulu kusukela ekushintsheni kwamandla kagesi amancane.
Ukubala kweCoulomb
Ukubala kwe-Coulomb, noma ukubala kuka-Ah, kuqagela isimo sokushaja ngokuhlanganisa ukuphuma nokuphuma kwamanje. Iqala ngesimo sokuqala senombolo yokushaja futhi iyishintshe njengoba kunyakaza kwamanje.
Isici Sokuhlola | Yekuchumana |
|---|---|
Indlela | I-algorithm yokubala ye-Coulomb ethuthukisiwe |
Indlela Yokuqinisekisa | Ukuhlolwa kwe-MATLAB uma kuqhathaniswa nesimo sangempela sokushaja kusuka kumajika wokushaja/wokukhipha |
Iphutha elikhulu (Ukuphela kokushaja) | Cishe 3.5% |
Iphutha Ngesikhathi se-CC Stage | Ngaphansi kuka 2% |
Iphutha Ngesikhathi Sesiteji Se-CV | Ngaphansi kuka 1% |
I-Trend yephutha | Iba nkulu ngokuhamba kwesikhathi ngaphambi kokuhlolwa kwezempilo |
Izinto Ezibalulekile | Isimo sokuqala esihle sokushaja kanye nokushaja kuhlola amaphutha aphansi |
Izinzuzo | Izibalo ezilula; ukunemba okuhle ngokwanele; ayikho idatha yebhethri eyengeziwe edingekayo |
Izithiyo | Amaphutha ayanda ngokuhamba kwesikhathi; idinga isimo sokuqala esihle sokukhokhiswa kanye nesimo sezinombolo zezempilo |
I-Coulomb Counting kulula ukuyisebenzisa futhi ayidingi idatha yebhethri eyengeziwe. Kodwa amaphutha angakhula ngokuhamba kwesikhathi. Amaphutha amancane esimweni samanje noma sokuqala senkokhelo angaba mabi kakhulu. Le ndlela isebenza kangcono ngokuhlola okuvamile noma ezinye izindlela zokusiza.
Indlela | I-RMSE | MSE | I-MFA | Ukutholwa Okusemqoka |
|---|---|---|---|---|
Ukubala kwe-Coulomb (CC) | 0.5071 | 0.2572 | 0.4571 | Amaphutha aphezulu kakhulu ngenxa yomsindo wezinzwa namaphutha; ayilungele ukusetshenziswa isikhathi eside |
Isihlungi Se-Kalman Esinwetshiwe | 0.0925 | N / A | N / A | Ukunemba okungcono ngosizo lwemodeli; idinga imodeli yebhethri enhle |
Ukunciphisa okukodwa | 0.0778 | N / A | N / A | Ingcono kune-EKF kodwa ayiphelele ekushintsheni kwesimo sezindleko |
Sekela Vector Machine | 0.0319 | N / A | N / A | Iphatha izinguquko kangcono; idinga amandla ekhompiyutha engeziwe |
Ukuhlehla Kwehlathi Okungahleliwe | 0.0229 | 0.0005 | 0.0139 | Ukunemba okungcono kakhulu; isebenza kahle ngomsindo nezinguquko; kuhle ekuphathweni kwebhethri langempela |
Ukuhlunga kwe-Kalman
I-Kalman Filtering isebenzisa amamodeli ezibalo ukuqagela isimo senkokhelo. Isihlungi esinwetshiwe se-Kalman nesihlungi se-Kalman esinganamakha sidumile. Lezi zihlungi zixuba idatha yesikhathi sangempela nokuqagela kwemodeli yebhethri. Balungisa ukuqagela kwabo njengoba idatha entsha ingena.
Izindlela zokuhlunga ze-Kalman ezifana ne-EKF, UKF, izihlungi ze-Kalman eziguquguqukayo, nezihlungi ezimbili ze-Kalman zisetshenziswa kakhulu.
Lezi zihlungi zisebenzisa amamodeli ebhethri alula namanye ayinkimbinkimbi ukuze zithole imiphumela engcono.
Ukuhlola kubonisa ukuthi izihlungi ze-Kalman zisingatha kahle izinguquko, inkumbulo yebhethri, nomsindo wezinzwa.
Ukushintsha izilungiselelo nokusebenzisa amanethiwekhi e-neural kuwenza abe ngcono nakakhulu.
Ukubuyekeza izinombolo kaningi kusiza ukulungisa amaphutha kusukela kuzinguquko zemodeli kanye nokushayela kwenzwa.
Ucwaningo lubonisa ukuthi izihlungi eziguquguqukayo nezikabili ze-Kalman zenza kangcono kune-EKF evamile esimweni sokukhokhiswa.
Ukuhlunga kwe-Kalman kunikeza ukuqagela okuhle, kwesikhathi sangempela sokushaja kwamabhethri e-lithium-ion. Idinga ukusethwa ngokucophelela kanye nemodeli yebhethri enhle. Kungaba nzima ukuyisebenzisa, kodwa isebenza kahle uma izinto zishintsha ngokushesha.
Izindlela ze-Hybrid ne-AI
Izindlela ze-Hybrid ne-AI zixuba izindlela ezisuselwe kumodeli nezindlela ezisuselwe kudatha ukuqagela isimo senkokhiso. Lezi zisebenzisa ukufundwa komshini, njengamanethiwekhi e-neural, ukusekela imishini ye-vector, nokuhlehla kwehlathi okungahleliwe. Bafunda kudatha ye-voltage, yamanje, kanye nezinga lokushisa. Izindlela zeHybrid zilungisa izinkinga indlela eyodwa engakwazi ngayo.
aspect | Incazelo |
|---|---|
Indlela | Isimo sokushaja esiyingxubevange sisebenzisa i-Coulomb Counting kanye Nomshini We-Vector Ehlobene (movIRVM-Coulomb) |
I-Dataset | Idatha yebhethri yebhethri eyodwa, idatha yokuhlolwa kwephakethe lebhethri, idatha yokulingisa ye-Advisor |
Izimo | Ukuhlolwa nge-US06, UDDS, NYCC, 1015 drive cycles; amazinga okushisa 0°C, 25°C, 45°C; isimo sokuqala sokukhokhisa 50%, 80% |
Ukunemba (RMSE) | Ngaphakathi kuka-2% ezivivinyweni eziningi namazinga okushisa |
Ngcono | Ngaphezulu kwama-30% angcono kune-movIRVM iyodwa; amaphutha ambalwa ngokuhamba kwesikhathi |
Isithiyo Esibalulekile Sixazululiwe | Ilungisa ukwakheka kwamaphutha ekubaleni okumsulwa kwe-Coulomb |
Amanothi engeziwe | Isebenzisa isilinganiso esinyakazayo ukusika umsindo; idinga kuphela idatha yokuqeqeshwa engu-10-30% yengxenye ye-RVM |
Izindlela ezihlanganisiwe zixuba idatha namamodeli ukuphatha izenzo zebhethri eziyinqaba.
Izindlela ezisuselwe kudatha zifaka amanethiwekhi e-neural, imishini yokusekela ye-vector, ukuhlehla kwenqubo ye-Gaussian, amanethiwekhi e-wavelet neural, kanye nokucabanga okungaqondakali.
Lezi zindlela ziqagela isimo sokushaja kumasiginali ongawalinganisa.
Izinkinga zihlanganisa umehluko webhethri, ukusetshenziswa okungajwayelekile, nokuguga kwebhethri.
Manje, abacwaningi bathanda izindlela ezisekelwe kudatha ngoba amamodeli ewodwa awakwazi ukulungisa zonke izinkinga.
Izifundo ezintsha ezisebenzisa ukufunda okujulile nedatha yemoto yangempela zibonisa izindlela ezixubile ne-AI zingaqagela isimo senkokhiso ngamaphutha angaphansi kuka-2%. Lezi zindlela zinembile kakhulu futhi zisebenza kahle, noma ngabe izinto zishintsha kakhulu.
Qaphela: Izindlela zezibalo zisiza isimo sokukhokhisa ukuqagela ngokulungisa ukungaqiniseki, amaphutha ezinzwa, nomsindo ongahleliwe. Ukulinganisa, ukwehla, nokuhlola kwenza zonke izindlela zokushaja zithembeke kakhulu.
Izindlela Zokulinganisa Isimo Sezempilo
Isimo sezempilo, noma i-SOH, isitshela ukuthi ibhethri le-lithium-ion seliguge kangakanani. Iqhathanisa ibhethri manje nesikhathi lapho lilisha. I-SOH itholakala ngokubheka umthamo wamanje futhi uyiqhathanise nomthamo wokuqala. Ingabuye ihlolwe ngokuqhathanisa ukumelana kwangaphakathi neseli entsha. Uma i-SOH yehla ngaphansi kuka-80% noma u-70%, ibhethri isekupheleni kokuphila kwayo. I-SOH ibalulekile ngoba ithinta indlela ibhethri elisebenza ngayo, ukuthi iphephe kangakanani, nokuthi ihlala isikhathi esingakanani. Njengoba i-SOH yehla, ibhethri libamba amandla amancane. Lokhu kusho ukuthi izimoto zikagesi azikwazi ukuya kude futhi imishini ayisebenzi isikhathi eside. Uma ibhethri liguga kakhulu, lingakhukhumala, livuze, noma libambe umlilo. Ukuqagela okuhle kwe-SOH kusiza ukumisa lezi zinkinga futhi kugcine amabhethri ephephile.
aspect | Ubufakazi | Idatha Yezinombolo / Imininingwane |
|---|---|---|
Incazelo ye-SOH | I-SOH isilinganiso somthamo wamanje kumthamo wokuqala noma iqhathanisa ukumelana kwangaphakathi nebhethri elisha. | Amazinga okuphela kwempilo e-SOH angamaphesenti angama-80 noma angama-70% omthamo osele. |
Umthelela Ekuphileni Okude | I-SOH ikhombisa ukuthi ungakanani umthamo olahlekile, okhawulela ukuthi izimoto zikagesi zingahamba ibanga elingakanani. Ukuguga kwebhethri kusho umthamo omncane. | Amabhethri emoto kagesi asetshenziswe ngaphezu kuka-10,000 km kanye nezinsuku ezingaphezu kuka-800 abonisa amaphethini okulahlekelwa amandla. |
Umthelela Ekuphepheni | Ukuguga okubi kungabangela ukuvuza, ukuvuvukala, ukushisa ngokweqile, nemililo. | Izingozi zokuphepha ziba zimbi kakhulu njengoba i-SOH yehla, ngakho ukuhlola i-SOH kubalulekile. |
Umthombo wedatha | Idatha ivela ezimotweni eziningi zikagesi ezinezindlela ezahlukene zokushayela nokushaja. | Idathasethi inezimoto zikagesi ezingu-347, amarekhodi ashajayo izinyanga ezingu-25, kanye nenqwaba yezinguquko zomhlaba wangempela. |
Izinselele ekulinganisweni kwe-SOH | Izinguquko zomhlaba wangempela, amaphutha ku-SOC, idatha enomsindo, kanye nokungatholakali kwamasampuli anele kwenza i-SOH ibe nzima ukuhlola. | Amaphutha e-SOC aba makhulu njengoba amabhethri ekhula, futhi i-BMS inenkinga yokubuyekeza umthamo ngokushesha. |
Izindlela Ezithuthukile | Ukufunda ngomshini nezindlela ezisuselwe kudatha zenza ukuhlola kwe-SOH kube ngcono. | I-BiGRU, ukwehliswa kwe-vector yokusekela, namanethiwekhi ajulile we-neural asiza ukuqagela i-SOH ne-SOC ngokunembile. |
Ukumelana Nangaphakathi
Ukumelana kwangaphakathi kubaluleke kakhulu ekuhloleni i-SOH kumabhethri e-lithium-ion. Njengoba amabhethri ekhula, ukumelana kwawo kwangaphakathi kuyenyuka. Lokhu kwenzeka ngoba izingxenye ezingaphakathi kwebhethri ziyaguga futhi ziwohloke. Uma ukumelana kuphindwe kabili noma umthamo wehla ufike ku-70-80%, ibhethri isekupheleni kokuphila kwayo. Izindlela eziningi zokuhlola i-SOH zisebenzisa ukumelana kwangaphakathi. Ukulinganisa ukumelana kunikeza ngokuqondile imiphumela emihle kodwa ngokuvamile kudinga ukuthi ibhethri liphumule, okunzima ngesikhathi sokusebenzisa okuvamile.
Ososayensi benze izindlela ezintsha zokusebenzisa ukumelana kwangaphakathi ukwenza ukuhlola kwe-SOH kube ngcono. Isibonelo, balungisa ijika le-voltage evulekile besebenzisa idatha yokumelana. Lokhu kusiza ukwehlisa amaphutha kusukela ekushintsheni kwesivinini sokushaja. Le ndlela isebenzisa izinto ezifana nesikhathi samanje sokushaja esikhundleni sezibalo eziqinile. Ukuhlolwa kwedatha yebhethri yangempela kubonisa ukuthi le ndlela ingehlisa iphutha eliphelele libe cishe ku-1.28% kwezinye izigaba zamandla kagesi. Le miphumela ibonisa ukuthi ukubuka ukumelana kwangaphakathi kwenza amasheke e-SOH aqine futhi anembe kakhulu.
Impedance
Izindlela ezisuselwe ekuphazamisekeni zisebenzisa indlela ibhethri elisabela ngayo kugesi ukuhlola i-SOH. Lezi zindlela zivame ukusebenzisa i-electrochemical impedance spectroscopy noma ukuhlolwa okufanayo. Ngokubona ukuthi ibhethri lisebenza kanjani ngamaza ahlukene, onjiniyela bangabona ukuguga futhi baqagele i-SOH. Izindlela ze-impedance zingaba neqiniso kakhulu, ngamaphutha e-root mean square phakathi kuka-0.75% no-1.5% wamayunithi we-SOH.
Uhlobo Lwendlela | Incazelo | Ukunemba kokubikezela kwe-SOH (Iphutha le-RMS) | Ukucatshangelwa Okusebenzayo |
|---|---|---|---|
Idatha ye-EIS eqondile | Isebenzisa idatha eluhlaza ye-electrochemical impedance spectroscopy | 0.75% - 1.5% amayunithi e-SOH | Ngokushesha ukulinganisa, kodwa amaseli angahluka |
I-Equivalent Circuit Fits | Ifanisa idatha ye-EIS namamodeli wesekethe | 0.75% - 1.5% amayunithi e-SOH | Idinga umsebenzi owengeziwe nezibalo, kodwa inokungaqiniseki okuncane |
Ukusabalalisa Kwezikhathi Zokuphumula (DRT) | Kubheka ukuthi kuthatha isikhathi esingakanani ukuthi izinto zilungiswe kusetshenziswa idatha ye-EIS | 0.75% - 1.5% amayunithi e-SOH | Kuthatha amandla amaningi ekhompyutha, kodwa kuyavumelana nezimo |
I-Nonlinear Frequency Response Analysis (NFRA) | Isebenzisa idatha yefrikhwensi ekhethekile ukuhlola i-SOH | 0.75% - 1.5% amayunithi e-SOH | Inikeza ulwazi oluhle mayelana nezenzo zebhethri, ngokushesha kunokukhishwa okugcwele |
Izindlela ezisuselwe ekuphazamisekeni zisebenza kahle kumalebhu futhi zinikeza imininingwane eminingi mayelana nokuguga kwebhethri. Kodwa lezi zindlela zingaba nzima futhi zibe nzima ukuzisebenzisa kumasistimu ebhethri esikhathi sangempela. Ngokuvamile badinga amathuluzi akhethekile nokusetha ngokucophelela. Izindlela ezintsha ezisuselwe kudatha seziqala ukuthatha izintambo ngokusebenzisa ukufunda komshini ukuqagela ukuguga kwebhethri ngaphandle kwamamodeli aqinile.
Ukubala Komjikelezo
Ukubala komjikelezo kungenye yezindlela ezindala kakhulu zokuhlola i-SOH kumabhethri e-lithium-ion. Le ndlela ibala ukuthi ibhethri ishajwe futhi isetshenziswe kangaki. Umjikelezo ngamunye ogcwele wenza ibhethri likhule kancane. Ngokubala imijikelezo, onjiniyela bangaqagela ukuthi ibhethri seliguge kangakanani.
Ukubala komjikelezo kulula futhi akudingi amathuluzi akhethekile noma izibalo eziqinile. Kodwa akubheki ukuthi umjikelezo ngamunye uhluke kanjani. Izinto ezifana nezinga lokushisa, ukuthi ibhethri lisetshenziswa malini, nokuthi lishaja ngokushesha kangakanani konke kushintsha ukuthi liguga kangakanani, kodwa ukubala komjikelezo kuphatha umjikelezo ngamunye ngendlela efanayo. Lokhu kungenza amasheke e-SOH angalungile, ikakhulukazi empilweni yangempela lapho amabhethri ebhekana nezinhlobo eziningi zokucindezeleka.
Izindlela Ezithuthukile
Izindlela ezithuthukisiwe zokuhlola i-SOH zisebenzisa ukufunda komshini kanye nobuhlakani bokwenziwa ukutadisha idatha yebhethri eminingi. Lezi zindlela zifunda ku-voltage, zamanje, nezinga lokushisa ukuqagela i-SOH kangcono kunezindlela zakudala. Amamodeli okufunda omshini afana nemishini ye-vector yokusekela, amahlathi angahleliwe, namanethiwekhi ajulile e-neural angathola amaphethini okuguga ebhethri akhohlisayo.
Ucwaningo lwakamuva lubonisa ukuthi lezi zindlela ezisekelwe kudatha zisebenza kangcono kunamamodeli angokwenyama amadala. Isibonelo, ukuhlehla kwevekhtha yokusekela kanye nokuhlehla kwenqubo ye-Gaussian kungathola impande yephutha eliyisikwele ngaphansi kuka-0.4% uma uqagela i-SOH. Amanethiwekhi emizwa yokuphakela phambili kanye nezinhlelo ze-adaptive neuro-fuzzy inference nazo enza kahle, ngamaphutha aphansi nemiphumela emihle yamabhethri ahlukene.
Izindlela zokufunda zomshini azidingi amamodeli ebhethri anemininingwane.
I-Cloud computing ivumela amamodeli amakhulu ukuthi asebenze, okwenza ukuhlola kwe-SOH kube ngcono ngisho noma isistimu yebhethri incane.
Ukusebenzisa imodeli yokufunda yomshini engaphezu kweyodwa kungenza ukuhlola kwe-SOH kucace nakakhulu.
Lezi zindlela zingathola amaphutha aphelele phakathi kuka-3% futhi zisuse amaphutha ayisikwele ngaphakathi kuka-2% ekuhlolweni kwangempela.
Kodwa, izindlela ezithuthukile zidinga idatha enhle neningi yokuqeqeshwa. Bangaba nenkinga ngokuguga kwebhethri okungajwayelekile noma izinguquko ezinkulu endleleni amabhethri asetshenziswa ngayo. Ukukhetha izici ezinhle ekushajweni kwedatha kubalulekile, njengoba ukushaja kuyinto evamile kunokusebenzisa ibhethri ezimotweni zikagesi. Onjiniyela kufanele baqinisekise ukuthi lezi zindlela ziqinile futhi ziphephile ngaphambi kokuzisebenzisa amasistimu ebhethri avikela abantu.
Qaphela: Ukusuka kumamodeli angokwenyama amadala ukuya ezindleleni ezisuselwe kudatha kubonisa ukuthi sidinga ukuhlola okungcono nokuvumelana nezimo kwe-SOH kwamabhethri e-lithium-ion. Ukufunda ngomshini kusiza ukubona ukuguga kwebhethri kusenesikhathi futhi kwenza amabhethri asebenze kangcono ngokuthola izimpawu zezinkinga ngokushesha.
Ukuhlanganisa Izindlela Zokunemba
Izindlela Ezihlanganisiwe
Amasistimu okuphatha amabhethri asebenza kangcono uma esebenzisa indlela engaphezu kweyodwa ukuhlola isimo sokushaja nesimo sempilo. Indlela eyodwa iyodwa ayikwazi ukuxazulula zonke izinkinga ezinhlelweni zebhethri ye-lithium-ion. Izindlela ezihlanganisiwe hlanganisa amandla asuselwa kumodeli, aqhutshwa yidatha, nokufunda. Lokhu kusiza ukunciphisa umsindo, ukuphatha okungaziwa, nokuhambisana nokuguga kwebhethri.
Ama-algorithms amaningi okuthuthukisa, njengezikwele okungenani, i-sunflower Optimization Algorithm, kanye ne-algorithm yokusesha ye-bald eagle, enza ukuhlola isimo sokushaja sibe ngcono. Isibonelo, i-algorithm yosesho lwe-bald eagle ibe nephutha eliphakeme elingu-1.06% kuphela le-SOC.
Amamephu Okuzihlela Athuthukisiwe nokufunda okugadwe kancane kubonise amaphutha aphezulu eduze kuka-1.25% kanye ne-RMSE ephansi ngo-0.55%. Le miphumela isho ukuthi izindlela ezixubile zinikeza amasheke aqinile e-SOC amabhethri e-lithium-ion.
Ukusebenzisa ukulinganisa kwamaseli okusebenzayo ngokufunda komshini ukuze uhlale uwusizo kusiza ngokuhlukana kwamaseli nokuguga kwebhethri. Amaseli abhalansi anikeza idatha yesimo esingcono sokushaja, esiza ukubikezela impilo yebhethri ye-lithium-ion.
Amamodeli enethiwekhi ye-hybrid neural network asiza ngokushintsha kwezinga lokushisa nokuthi amabhethri asetshenziswa kanjani. Ngokuxuba ukulinganisa ngokomzimba nezindlela eziqhutshwa idatha, amasistimu okuphatha ibhethri angasiza amabhethri e-lithium-ion ukuthi ahlale isikhathi eside futhi asebenze kangcono. Ukuhlanganiswa kwamamodeli amaningi, njengeHlathi elingahleliwe, kwenza ukuhlolwa kwezempilo kuqine nakakhulu ngokusebenzisa izingxenye ezingcono kakhulu zamamodeli ahlukene.
Izindlela ezihlanganisiwe zisiza amasistimu okuphatha ibhethri aphathe izinguquko zomhlaba wangempela. Lokhu kuzenza zithembeke kakhulu ezimotweni zikagesi nokunye ukusetshenziswa.
Ukucatshangelwa Kokusebenza
Ukukhetha nokusebenzisa izindlela ezixubile ezinhlelweni zangempela zebhethri ye-lithium-ion kudinga ukuhlela ngokucophelela. Onjiniyela kufanele bacabange ngalokho okudingwa ukusetshenziswa ngakunye, njengezimoto zikagesi noma indawo yokugcina izinto.
Izindlela eziqhutshwa idatha zisebenzisa idatha yenzwa yesikhathi sangempela futhi ziyashintsha njengoba amabhethri ekhula noma asetshenziswa. Lezi zindlela zinembe kakhulu, zisebenza ngamakhemistri ahlukene, futhi ziwuphatha kahle umsindo wezinzwa.
Ama-Hybrid frameworks axuba ama-algorithms angcono wehlathi angahleliwe, amamodeli asuselwa ku-physics, namanye amathuluzi okufunda omshini. Le bhalansi inikeza ukunemba, isebenza ngokushesha, futhi ingasetshenziselwa izinhlobo eziningi zebhethri ye-lithium-ion nezimo.
Onjiniyela kufanele baxazulule izinkinga ezinjengokudinga idatha eningi enhle, ukukhetha izici ezifanele, nezindleko zekhompyutha. Ukuxuba izici nezilungiselelo zokushuna kungenza ukuqagela kube ngcono futhi kusize ngezinguquko zesikhathi sangempela.
Idatha eminingi, njenge-voltage yeselula, yamanje, izinga lokushisa, nesibalo somjikelezo, isiza ukukhetha izindlela ezingcono kakhulu zengxubevange. Lezi zindlela zisiza ngedatha enomsindo noma engekho futhi zinikeze imiphumela ekhethekile yokusetshenziswa ngakunye, hhayi nje isimo esiyisisekelo sezindleko nesimo sempilo. Empilweni yangempela, izindlela ezixubile zisebenza kahle kumalebhu nasensimini, njengasezimotweni zikagesi, lapho zigcina amabhethri ephephile futhi esebenza ngaphansi kwezimo ezihlukene.
Ithiphu: Uma ukhetha izindlela ezixubile, onjiniyela kufanele bafanise indlela nemigomo yesistimu yebhethri, idatha, nalapho izosetshenziswa khona. Lokhu kusiza ukuqinisekisa ukuthi ukuphathwa kwebhethri le-lithium-ion kuthembekile, kungakhula, futhi kusebenza ngesikhathi sangempela.
Ukwazi i-soc elungile ne-SOH kubaluleke kakhulu ekutheni amabhethri e-lithium-ion asebenza kahle futhi aphephe kangakanani. Yonke indlela inamaphuzu ayo amahle, kodwa ukusebenzisa indlela engaphezu kweyodwa ndawonye ohlelweni lokuphatha ibhethri kunikeza imiphumela engcono kakhulu yokwenza amabhethri e-lithium-ion agcine futhi asebenze kangcono. Ucwaningo olusha lubonisa ukuthi ukusebenzisa izindlela ezihlakaniphile zokukhetha idatha ebalulekile kanye namanethiwekhi athuthukisiwe e-neural angenza amaphutha abe mancane kakhulu, ngisho ehle aze afike ku-0.16%. Lokhu kusiza amabhethri ukuthi ahlale isikhathi eside futhi ahlale ephephile. Kubalulekile ukukhetha indlela yokulinganisa elingana nalokho okudingwa yibhethri ye-lithium-ion ngayinye.
Imibuzo Evame Ukubuzwa
Uyini umsebenzi oyinhloko wesistimu yokuphatha ibhethri?
Isistimu yokuphatha ibhethri igcina amabhethri ephephile. Ihlola isimo senkokhelo kanye nesimo sempilo. Uhlelo lubhalansa amaseli ukuze asebenze ndawonye. Ivimba amabhethri ukuthi ashise kakhulu noma agcwale kakhulu. Lokhu kusiza amabhethri ukuthi ahlale isikhathi eside futhi asebenze kangcono.
Kungani izinzwa zingakwazi ukukala isimo sokushaja ngokuqondile?
Izinzwa azikwazi ukubheka ngaphakathi kwebhethri. Ukusabela kwamakhemikhali kwenzeka ngaphakathi lapho izinzwa zingakwazi ukubona. Izinzwa zikala kuphela i-voltage, yamanje, nezinga lokushisa. Isistimu isebenzisa lezi zinombolo ezinama-algorithms akhethekile ukuqagela isimo senkokhelo.
Izinga lokushisa likuthinta kanjani ukulinganisa kwesimo sebhethri?
Uma kushisa kakhulu noma kubanda, ukusabela kwebhethri kuyashintsha. Isistimu ingase yenze amaphutha esimweni sokukhokha noma isimo sempilo. Amasistimu okuphatha amabhethri amahle ashintsha izibalo zawo ukuze alungise lawa maphutha.
Iyiphi indlela enikeza isilinganiso esinembe kakhulu sesimo sezempilo?
Indlela | Ukunemba Izinga |
|---|---|
Ukufunda komshini | Phezulu kakhulu |
Ukuhlaziywa kwe-Impedans | High |
Ukumelana Nangaphakathi | Medium |
Ukubala Komjikelezo | ongaphakeme |
Ukufunda ngomshini ngokuvamile kunikeza imiphumela engcono kakhulu uma idatha iyinhle.
