Ukuhlanganiswa kwe-AI kanye ne-IoT ezinhlelweni Zokuhlanza Irabha

By UMike Chen, Umqondisi Wokukhiqiza | Iminyaka eyi-12+ Ekukhiqizeni Irabha |I-LinkedIn

Izinto Ezibalulekile Okufanele Uzicabangele

• Umgqomo owodwa we-XCJ-G600 ujikeleza ibhetshi engu-15 kg njalo ngemizuzwana engu-20-40 — imijikelezo engaphezu kuka-1,300 ecebile ngedatha ngokushintsha kwamahora angu-8 isivele itholakala ukuze ithathwe yi-IoT.
• Amayunithi okususa ukungcola e-cryogenic asebenza emazingeni okushisa afinyelela ku-−150 °C ngesivinini sama-roller esingu-20-70 RPM, futhi ukubhala phansi kokubili okuguqukayo kuveza ngqo imfucuza ye-nitrogen kanye nokukhukhuleka kokushisa.
• Ngenxa yokuthi imishini exhunyiwe yembula ukushelela kwenqubo kusenesikhathi, ukulungiswa kokubikezela kuvame ukuguqula ukumiswa okungahleliwe kube ukungenelela okuhleliwe okungaphansi kwemizuzu engama-30 (amanani ajwayelekile avela kumaphrojekthi amakhasimende ethu).
• Ukushintsha ama-beats: imishini eqhutshwa yi-inverter efana ne-7.5 kW XCJ-G600 ingaxhunywa ngezinzwa kanye nesango elingaseceleni esikhundleni sokuthenga okusha.

Iningi labakhiqizi esisebenza nabo kakade banemishini yokususa ukungcola ehlakaniphile ezingeni lomshini futhi ethule ngokuphelele ezingeni lefektri. Isilawuli siyazi isivinini se-roller; umqhubi uyazi isisindo se-batch; ithimba lekhwalithi liyazi izinga lokwenqatshwa - kodwa akekho oxhumanisa lokhu okuthathu. Lokho kunqanyulwa yilokho kanye amaphrojekthi e-AI ne-IoT aklanyelwe ukukulungisa.

Lesi sihloko sichaza ukuthi ukuhlanganiswa kusho ukuthini ngempela esimweni sokunciphisa ijubane, ukuthi yimaphi amaphuzu edatha okufanele ahlanganiswe kuqala, kanye nohlelo oluhlelwe ngezigaba olusebenza ezitshalweni zerabha ezincane neziphakathi nendawo kunezimboni ezihlakaniphile ezisezingeni eliphezulu.

Kusho ukuthini ngempela i-AI ne-IoT kumongo wokunciphisa ijubane?

I-Internet of Things (IoT) iwumkhuba wokuxhuma imishini yezimboni kunethiwekhi ukuze idatha yayo yokusebenza iqoqwe, igcinwe, futhi ibuyekezwe; ubuhlakani bokwenziwa (AI) ungqimba lwesofthiwe oluthola amaphethini kuleyo datha futhi luncome noma lufeze izinqumo.Ekuguquleni i-deflashing, lokhu kusebenza njengesibili: i-IoT inikeza isignali eluhlaza, kanti i-AI iguqula isignali ibe yisenzo - yehlisa i-roller, ifaka uphawu ebhandeni eligugile, noma ihlukanise i-batch esolwayo ngaphambi kokuba ithunyelwe.

Kuyasiza ukucabanga ngomshini wokukhipha ugesi njengomthombo wedatha kuqala bese kuba ithuluzi lokususa ugesi okwesibili. Umjikelezo ngamunye wokukhipha ugesi ngomshini kanye nokugijima ngakunye okungaphansi kuka-−120 °C kukhiqiza umunwe olandelekayo wejubane, isikhathi, izinga lokushisa, kanye nomthwalo. Uma lezo zigxivizo zeminwe sezigciniwe, imibuzo eyayidinga unjiniyela wenqubo emi emgqonyeni ukuze asebenze ngokugcwele — “izikhathi zomjikelezo ziqale nini ukwanda?” — iba umbuzo wedeshibhodi wemizuzwana eyishumi.

Yimaphi Amaphuzu Edatha Okususa Imfucuza Afanele Ukuxhumaniswa Kuqala?

Qala ngeziguquguquko ezine esezinqumile kakade ikhwalithi yokususa ukungcola: isikhathi somjikelezo, isivinini se-roller, izinga lokushisa lenqubo, kanye nomthwalo we-batch.Zikhona kuwo wonke umshini, ziyashintsha uma kukhona okungahambi kahle, futhi zihambisana ngqo nemiphumela yomkhiqizo. Ithebula elingezansi lihlanganisa i-variable ngayinye namanani angempela emishini kanye nokwenza ngcono okuvulayo.

Amanani emishini atholakala ku-Ikhasi lomkhiqizo le-XCJ-G600kanyeimininingwane yomshini wokuhlanza i-nitrogen cryogenic owuketshezi.

Ngenxa yokuthi i-cryogenic embrittlement iwumjaho ophakathi kwezinga lokushisa nesikhathi, ama-logged cooling curves ayisethi yedatha ebaluleke kakhulu emgqeni we-cryogenic.Ukugijima okufinyelela ekuvinjelweni okuhlosiwe ngemizuzu emi-5 esikhundleni sika-8 kusindise u-37% womjikelezo walo — ukonga okungabonakali kuwashi lokumisa kodwa okusobala kudatha.

Ukulungiswa Kwezinto Ezibikezelayo Kushintsha Kanjani Umnotho Owohlokayo?

Ukulungiswa kokubikezela kusebenzisa idatha yethrendi — izikhathi zomjikelezo ezikhuphukayo, ukushintshashintsha kwamanje kwenjini, ukungazinzi kwezinga lokushisa — ukuhlela isevisi ngaphambi kokuba ingxenye yehluleke, esikhundleni se-after.Umehluko wezomnotho ukhulu kakhulu: injini yokushayela ehlulekile phakathi nesikhathi sokusebenza iyamisa umugqa, iphazamise ishejuli, futhi ingasusa yonke inqwaba yemishini esetshenziswayo, kuyilapho ukushintshana okuhleliwe okubonwe kusasele amasonto amabili kwenzeka phakathi kokushintshana.

Ngokwesipiliyoni sethu sokusekela imigqa yokususa ukungcola ezitshalweni zamakhasimende, ukuqapha okuxhunyiwe kuvame ukuveza ukugeleza kwenqubo okusebenzayo okuthathu kuya kokuhlanu ngekota emshinini owodwa - okuvame kakhulu ukunwetshwa kwesikhathi somjikelezo kancane kancane kanye nokupholisa okungaguquki - futhi ngayinye ingalungiswa efasiteleni elihleliwe ngaphansi kwemizuzu engama-30 (amanani ajwayelekile avela kumaphrojekthi ethu amakhasimende; izibalo eziqondile ziyahlukahluka ngenhlanganisela ehlanganisiwe kanye nephethini yokushintsha).Ngenxa yokuthi i-inverter engu-7.5 kW emshinini ofana ne-XCJ-G600 isivele ibika umthwalo wemoto ngedijithali, izindleko zokungena kuleli zinga lokuqapha ziyisango kanye nokucushwa, hhayi ihadiwe entsha.

Kukhona ne-engeli yokuthobela imithetho. Izitshalo eziqinisekiswe ukuthi ziyasebenzaI-ISO 9001bacelwa ukuba babonise intuthuko eqhubekayo; umkhondo wedatha wamapharamitha omshini ngeqoqo ngalinye ungomunye wobufakazi obuhlanzekile kakhulu umhloli wezimali azobumukela.

Umbono Womshini Ungena Kuphi Ku-Quality Loop?

Umbono womshini uvala iluphu eshiywa yidatha yesenzwa iyodwa: amakhamera ahlola izingxenye eziphukile, bese ungqimba lwe-AI lulandelela yonke ingxenye enqatshiwe emuva kumjikelezo oqondile, isivinini, kanye nokushisa okwayikhiqiza.Lokhu kuguqula ukulawulwa kwekhwalithi kusuka emsebenzini wokubuka ngemuva kube okokufaka kwesiteringi.

Iketanga elisebenzayo libukeka kanje: izingxenye ziphuma emgqonyeni okhipha amanzi, zidlula ekuhlolweni kombono ukuthi i-flag residue iyakhanya yini, bese izingxenye ezihlanzekile ziqhubekaumshini wokuhlukanisa irabhaehlunga umkhiqizo omuhle kusuka kudoti wokusika. Lapho uhlelo lokubona lubhala iqoqo lokwenqatshwa, i-AI iqhathanisa izitembu zesikhathi nedatha yemigqomo bese iphendula umbuzo obalulekile: ingabe lokhu kwakuyikhompawundi embi, isivinini sokugoqa esingalungile, noma ibhetshi elalingama-2 kg ngaphezu kwe-spec?

Imigqa ye-Cryogenic izuza kakhulu kulokhu kubhangqa. I-Cryogenic deflashing ifinyelela amazinga okudlula komkhiqizo oqediwe angaphezu kwama-98% lapho amapharamitha ezinzile, ngakho-ke ngisho nokuphambuka okuncane okulotshiwe — ijika lokupholisa elisebenza ngo-10 °C elifudumele kune-baseline — kufanelekile ukubamba ngaphambi kokuba libize amaphuzu amathathu okukhiqiza.

Ibukeka Kanjani Imephu Yokuhlanganisa Engokoqobo?

Indlela esebenzayo yezitshalo eziningi zerabha iyizigaba ezintathu: ukuxhuma, ukuhlaziya, bese wenza ngcono — ngaleyo ndlela, emshinini owodwa ngaphambi komugqa wonke.Ukuzama ukwenza izinqumo ngokuzenzakalela ngaphambi kokuba idatha ibe khona kuyindlela evame kakhulu yokuhluleka esiyibona kumaphrojekthi e-Industry 4.0.

1. Xhuma umshini owodwa bese uchaza isisekelo.Faka isango elingaseceleni kuyunithi eyodwa yokususa ugesi — entsha kakhulu, eqhutshwa kakhulu yi-inverter — bese ubhala isikhathi somjikelezo, isivinini, izinga lokushisa, kanye nomthwalo izinsuku ezingu-60 ngaphandle kokushintsha noma yini, ngakho-ke "okuvamile" kulinganiswa kunokuthi kucatshangelwe.
2. Hlaziya uqhathanise nemiphumela yekhwalithi.Hlanganisa idatha yomshini namarekhodi okwenqatshwa avela ekuhlolweni, bese uvumela ungqimba lwe-AI (noma ispredishithi esihle, ngokweqiniso) lufune ukuhlobana phakathi kwemicimbi yokukhukhuleka kwamapharamitha kanye nekhwalithi.
3. Lungiselela ngezaziso ze-close-loop.Setha imikhawulo yesaziso kusukela ekuqaleni, uyiqondise kumafoni opharetha noma kudeshibhodi yesitshalo, bese ucabanga ngokulungiswa kwepharamitha okuzenzakalelayo - isinyathelo esidinga ukulawulwa okukhulu kanye nokungaphuthumi kakhulu.

Yiziphi Izilinganiso Eziqotho?

I-AI ne-IoT ngeke zilungise inqubo engazinzile, umgqomo ogugile, noma isikhunta esichazwe kabi - zizomane zibike lezo zinkinga ngokushesha nangokunembile.Ukuhlanganiswa kwandisa ukulawulwa kwenqubo okuhle; akukuthathi indawo yakho.

Ezinye izexwayiso ezimbili ezivela kokuhlangenwe nakho kwethu kwephrojekthi. Okokuqala, ukusebenza okuncane okuhlanganisiwe kukhiqiza idatha encane: imodeli ye-AI eqeqeshwe ngemijikelezo engu-300 enosayizi we-O-ring eyodwa ngeke idluliselwe kahle emndenini wezingxenye ezingu-40 zesikhathi esifushane, ngakho-ke okulindelwe kumele kubekwe ngokuhlanganiswa komkhiqizo ngamunye. Okwesibili, ukuxhumana kunobuso bokuphepha - isiqondiso njengezinsiza zokuphepha kwe-cyber zokukhiqiza ezishicilelwe yi-I-NISTkufanelekile ukufundwa ngaphambi kokuba isango liqale ukusebenza.Ngenxa yokuthi iseli elikhipha i-deflashing alivamile ukubhekwa yi-inthanethi ngokuklama, inethiwekhi yefektri ehlukaniswe kahle igcina ingozi eyengeziwe iphansi, kodwa "ephansi" akuyona "zero" futhi kufanele ibekelwe isabelomali sayo.

Isiphetho: Xhuma idatha osuvele unayo

Ukususa irabha kufaneleka kakhulu ekuhlanganisweni kwe-AI kanye ne-IoT, ngesizathu esisodwa esilula: imishini isivele isebenza eziguquguqukweni ezilawulwa ngokunembile nezilinganiswayo. Ibhetshi engu-15 kg, umjikelezo wemizuzwana engu-20-40, uphahla lwe-−150 °C, ububanzi be-roller engu-20-70 RPM — lawa amaphuzu edatha alinde ukuqoqwa, futhi wonke awo aphethe ulwazi lwekhwalithi.

Uma uhlela ukutshalwa kwezimali kwemishini elandelayo, khetha imishini enabalawuli abadalula leyo datha ngendlela ehlanzekile.uhla oluphelele lomshini wokuhlanza irabhakusuka ku-Xiamen Xingchangjia, noma usithumelele imidwebo yakho yezingxenye kanye nomphumela oqondiwe — ithimba lethu lobunjiniyela lizophakamisa ukucushwa kokususa ugesi ngezinketho zokuxhuma ezihambisana nezinga lokuzenzakalela kwesitshalo sakho.