Binciken Bayanai na Ƙarfe da Ƙarfi na AI tare da Python

Riba:

• Ikon samarwa da fahimtar lambar da ke tsaftacewa, tantancewa da hango bayanan ƙarfe tare da Python, tare da tallafin AI
• Ikon kafa al'ada ta tabbatar da lambar AI da aka ƙirƙira tare da duba naúrar, shari'ar gwaji da ƙwarewar jiki.
• Cikakken fahimta na iyakoki na ƙarshe zuwa ƙarshe, alhakin da tsarin ɗabi'a na AI a cikin ƙarfe

Ana ƙara yin aikin ƙarfe tare da ƙarin bayanai: ɗaruruwan sakamakon gwajin tensile, dubban ma'auni na SPC, bayanan tanderun maganin zafi, abubuwan da aka nuna. Neman wannan bayanan da hannu a cikin Excel yana da jinkirin da kuskure; Python (harshen shirye-shirye na gama gari don nazarin bayanai da ƙididdiga na kimiyya) ya sa wannan maimaituwa, abin dubawa, da atomatik. AI yana da ƙarfi sosai a rubuce-rubuce, gyarawa, da bayyana lambar Python - amma lambar da take samarwa na iya yin daidai amma ba da sakamakon da ba daidai ba: kuskuren naúrar, ƙima mara kyau, asarar bayanan shiru. A cikin wannan rukunin na ƙarshe, zaku koyi yadda ake bincika bayanan ƙarfe tare da tallafin AI a cikin Python kuma ku tabbatar da kowace lamba, sannan ku cika iyakokin AI, alhakin da tsarin ɗa'a wanda aka rufe cikin duka tsarin.

Ayyuka na yau da kullun a cikin nazarin bayanan ƙarfe

Ayyukan gama gari tare da tallafin AI a cikin Python:

• Tsabtace bayanai: Ƙimar da ta ɓace, rudani na ɗaya (MPa vs. ksi), cirewa daga waje. Bayanan datti yana lalata har ma mafi kyawun bincike.
• Kallon gani: Zana lanƙwan ƙarfi, ginshiƙi mai sarrafawa, rarraba girman hatsi, bayanin martaba-lokaci.
• Hesap ve ilişki: Hall-Petch (tane boyutu-akma dayanımı ilişkisi: σ = σ₀ + k·d^(-1/2)), Arrhenius (sıcaklık-hız), regresyon, istatistik.
• Automation: Batch sarrafa ɗimbin fayiloli da samar da tebur / rahotanni na taƙaitaccen bayani.

Dangantakar Hall-Petch misali ne mai kyau a nan: ƙarfin yawan amfanin ƙasa yana ƙaruwa daidai da tushen murabba'in diamita na hatsi. AI na iya fassara wannan dabara zuwa lamba; Amma ko kuna amfani da d azaman micrometer ko mita kuma wace naúrar da kuke amfani da ita don k coefficient gaba ɗaya yana canza sakamakon. Daidaiton lambar ba yana cikin haddar dabarar ba, amma a cikin tabbatar da raka'a da shari'ar gwaji.

Tukwici: Tabbatar cewa kun haɗa da "sanannen gwajin amsa" a cikin kowane asusun da AI ke rubutawa. Ba da ƙaramin shigar da sakamakon da kuka sani da hannu (misali ƙarfin da ake tsammani don d = 100 µm); Idan lambar ta ba da wannan ƙimar, za ku iya amincewa da shi; idan ba haka ba, kuskuren yana ɓoye.

Mataki-mataki: Amintaccen bincike yana gudana tare da AI

1. Bayyana aikin da bayanai: Wane fayil, waɗanne ginshiƙai, waɗanne raka'a, menene za a ƙididdige su?
2. Neman kwarangwal na lamba: Tambayi AI don yin sharhi, kwarangwal da aka raba mataki-mataki; Ƙayyade raka'a tare da bayyanannun masu canji a saman.
3. Gwaji tare da ƙananan bayanai: Gudu tare da ƴan layuka / sanannun dabi'u na farko; Duba sakamakon da hannu.
4. Naúra da sarrafa iyaka: Shin abin da aka fitar a zahiri yana da ma'ana? Shin akwai wani maganar banza kamar ƙarfi mara kyau, ƙimar 120%?
5. Yi hangen nesa da duba: Shin jadawali yana nuna yanayin da ake tsammani? Shin kurakuran bayanan da suka fita waje?
6. Tabbatar da daftarin aiki: Kwatanta sakamakon zuwa lissafi/misali mai zaman kansa; Yi sharhi da code da zato.

Teburin tabbatar da lambar AI

Hadarin

alama

Hanyar sarrafawa

Kuskuren naúrar

Sakamakon shine 1000x karkatacciyar hanya

Ƙayyade raka'a a madaidaicin madaidaici, sarrafa girman

Ƙididdigar ƙididdiga / tsari mara daidai

Ma'ana amma ƙima mara kyau

Gwaji yanayin tare da sanannun amsoshi

Asarar bayanan shiru

Adadin layuka ya yi ƙasa da yadda ake tsammani

Buga lambar layuka kafin/bayan aiki

m karkacewa

matsakaicin kayak

Yi hangen nesa tare da jadawali, duba rarraba

Overfitting (laburare na hallucinatory)

aikin da ba shi da shi

Gudu, karanta kuskure, duba takardu

uku mini lokuta

Case 1 - Kuskuren ƙarar ɓoye. Injiniya yana buƙatar lamba daga AI don lissafin Hall-Petch; Lambar tana tsammanin diamita na hatsi a cikin mita, amma injiniyan ya shiga micrometers. Sakamakon shine sau 1000 karkatacciyar hanya, amma ba a iya gani a farkon kallo saboda yana kama da "lambar". Mühendis d = 100 µm için elle beklediği ~350 MPa ile kodun verdiği absürt değeri karşılaştırınca hatayı yakalar, birim dönüşümünü koda ekler. Darasi: Gwajin amsa da aka sani yana bayyana kuskuren naúrar a cikin daƙiƙa.

Case 2 - Silent data asarar. Ƙungiya tana tsaftace sakamakon gwajin ja na 500 tare da lambar AI da aka rubuta; Yayin da lambar ke sauke layuka tare da bacewar ƙima, kuma ta bazata ta watsar da wasu layuka masu inganci, suna barin layuka 300. Matsakaicin ƙarfin yana da girma fiye da yadda ake tsammani saboda an kawar da ƙarancin ƙima. Lokacin da ƙungiyar ta buga adadin layuka kafin/bayan aiki, sun ga cewa layuka 200 sun ɓace cikin shiru kuma suna gyara tacewa. Darasi: Buga adadin layuka a kowane matakin sarrafa bayanai yana kama da asarar shiru.

Case 3 - Daidaitaccen amfani, rahoto mai maimaitawa. Ƙungiya mai inganci dole ne ta aiwatar da dumbin bayanan tanderun da kuma samar da rahotanni taƙaice kowane mako. Yana tambayar AI don rubuta rubutun da aka yi sharhi wanda ke karanta fayiloli da yawa kuma yana samar da hotuna da tebur na taƙaitaccen bayani. Ƙungiyar tana gwada lambar tare da wasu sanannun fayiloli, tabbatar da adadin raka'a da layuka, suna duban jadawali na gani. Sannan yana amfani da rubutun kowane mako; Ana rage sa'o'i na aiki zuwa mintuna kuma suna aiki iri ɗaya, hanyar dubawa kowane mako. AI ya ba da damar maimaita aiki da kai, ƙungiyar ta tabbatar.

Samfuran gaggawar da za a iya kwafi

CODE SKELETON TEMPLATE"Rawar: Kai ne mataimaki na lambar don nazarin bayanan ƙarfe. Aiki: [misali karanta sakamakon zana daga CSV da matsakaicin ƙarfin yawan amfanin ƙasa]. Bayanai: [fayil, ginshiƙai, UNITS]. rubuta kwarangwal da aka yi sharhi tare da pandas: (1) ayyana raka'a a cikin buɗaɗɗen mabambanta a saman, (2) karanta bayanai da ƙididdige jeri na 4, (2) karanta bayanai da ƙidaya a layi (3 jere) lissafi, (5) buga sakamakon da kuma duba yiwuwar ƙara sharhi zuwa kowane mataki Alama inda nake buƙatar tabbatarwa.

KNOWN AMSAR TEMPLATE"Rubuta 'sanannen gwajin amsawa' don aikin mai zuwa: [aiki]. Zaɓi ƙaramin shigarwa wanda za'a iya ƙididdige sakamakonsa da hannu, yi sharhi abin da ake sa ran, sannan a duba tare da tabbatarwa. Hakanan gwada daidaiton naúrar. Idan gwajin bai wuce ba, jera dalilai masu yiwuwa (unit, coefficient, formula)."

KYAUTATA KYAUTA / DEBUG TEMPLATE"Bincika lambar da ke gaba: [lambar]. Bincika kuma tuta mai zuwa: (1) daidaiton ƙara, (2) haɗarin asarar bayanan shiru, (3) tasiri mai ƙarfi, (4) fitowar da ba za ta iya yiwuwa ba (ƙarfin mara kyau,> ƙimar 100%). Rubuta yadda ake tabbatar da kowane haɗari.

KYAUTA TEMPLATE "Plot [tensile curve / control plot / rarraba] tare da matplotlib. Ƙara UNITS zuwa alamomin axis, sanya lakabi da grid. Yi alama daban-daban. Lura cewa ina buƙatar dubawa na gani da kuma tabbatar da yanayin da ake tsammani (misali ƙara yawan damuwa). Kada ku ɓoye kowane maki don 'kawata' bayanan. "

Rauni mai ƙarfi / Ƙarfi mai ƙarfi

RAUNI GASKIYA: "Rubuta min lambar da ke yin lissafin Hall-Petch."

KARFI GASKIYA:"Rawar: Kai ne mataimakin lambar don nazarin bayanan ƙarfe. Rubuta aikin Python don ƙididdige Hall-Petch(σy = σ0 + k·d^(-1/2)) Raka'a: d micrometer, σ0 da σy MPa, k MPa·µm^(1/2) = Juya juzu'in naúrar ON a cikin aikin da kuma sharhi 1 . µm,σ0 = 50 MPa, k = 300 MPa·µm^0.5 kuma gwada tare da tabbatarwa.

Rarraunan hanzari yana samar da lambar da ba ta da naúra da gwaji kuma yana iya yin kuskure cikin shiru. Ƙarfi mai ƙarfi yana bayyana raka'a a sarari, tilasta jujjuyawa, yana buƙatar sanannun gwajin amsa, da kuma tantancewa - mai tabbatar da lambar.

Cikakken taƙaitaccen tsarin: Wuri da iyakokin AI a cikin ƙarfe

An sake maimaita ƙa'ida ɗaya a cikin wannan tsarin: AI mataimaki ne, ba mai yanke shawara ba. A cikin raka'a goma sha biyu, mun yi amfani da AI daga ƙarshen zuwa-ƙarshe, daga zaɓin kayan abu zuwa tsinkayar dukiya, nazarin lokaci zuwa microstructure, maganin zafi zuwa ƙirƙira, kula da inganci don lalata bincike, haɓakawa zuwa daidaitawa da bincike na Python. A cikin kowane rukunin, AI ya bayyana ra'ayoyi, ya haifar da hasashe, ya rubuta zayyana da lamba - amma a cikin kowane ɗayan da bai yanke shawara shi kaɗai ba. Iyakokin sun bayyana a sarari: hallucination (lambar ƙira da ƙa'ida), rashin wakilcin alhakin aminci-muhimmanci, dogaro ga ingancin bayanai da ma'auni/ daidaitawa, haɗarin ɓarna da sirri. Horon tabbatarwa ya kasance koyaushe tare da mutane a cikin ma'auni-laboratory-asusun alwatika.

Kuskuren gama gari

• Aiwatar da lambar zuwa ainihin bayanai ba tare da gwaji tare da sanannun amsoshi ba.
• Rashin bayyana raka'a a sarari; Duban MPa/ksi, µm/m rudani.
• Ba buga adadin layuka yayin matakan sarrafa bayanai ba kuma rashin lura da asarar shiru.
• Dogaro da ma'ana ba tare da hangen nesa ba.
• Karɓar ɗakin karatu / aikin da AI ba ya wanzu ba tare da aiwatar da shi ba.
• Tsammanin lambar daidai ne saboda "da alama yana aiki"; gazawar duba plausibility na jiki.

A takaice

AI a cikin binciken bayanan ƙarfe tare da Python babban mai haɓakawa ne don ƙididdigewa, cirewa, da sarrafa kansa; amma lambar da yake samarwa na iya ɗaukar kurakuran naúrar, ƙayyadaddun ƙididdiga marasa daidai, da asarar bayanan shiru. Tabbatar da kowace lamba tare da sanannun gwajin amsa, duba raka'a, bin diddigin layuka, da yuwuwar zahiri. Anan, kamar yadda yake a cikin dukkan nau'ikan, AI shine mataimaki: yana tabbatar da saurin gudu da tsari, injiniyan da aka ba da izini ya karɓi daidaito da alhakin. A cikin aminci-mahimmancin yanke shawara na ƙarfe, fitowar AI ba ta taɓa zama madadin ma'auni, gwaje-gwaje, ƙididdiga da ƙwararriyar amincewar injiniya ba.

Aikin aikace-aikace

Zaɓi lissafin ƙarfe na ƙarfe (misali Hall-Petch ko matsakaicin bayanan tensile). Tare da samfurin "CODE SKELETON", tambayi AI don lambar sharhi wanda ke buga adadin layi tare da raka'a a sarari. Sannan ƙara yanayin gwaji inda kuka san sakamakon da hannu tare da samfurin "SANAN ANSWER TEST" kuma kunna shi. Idan lambar ba ta ba da ƙimar da ake tsammani ba, nemo kuskuren (naúrar, ƙididdiga, bayanai) kuma gyara shi. A ƙarshe, rubuta game da horo na tabbatar da AI da kuka koya a cikin wannan rukunin ta hanyar daidaita shi zuwa kasuwancin ku a cikin labarai uku.

jerin abubuwan dubawa

• [] Na bayyana a sarari aiki, bayanai da raka'a zuwa AI.
• [ ] Na ayyana raka'a tare da bayyanannun masu canji a saman lambar.
• [ ] A cikin matakan sarrafa bayanai, na buga adadin layuka kuma na duba asarar shiru.
• [ ] Na tabbatar da lambar tare da shari'ar gwaji tare da sanannun amsa.
• [ ] Na duba iyawar zahiri (mara kyau/marasa ƙima) na fitarwa.
• [] Na yi amfani da AI a matsayin mai sauri; Na kiyaye mutunci da alhaki.

Gwajin Module

1. Lokacin aiki tare da kayan aikin AI a matsayin injiniyan kayan aiki, babban alhakin wanne daga cikin waɗannan yakamata ya kasance koyaushe tare da ɗan adam ( injiniyan doka)?

• A) Amincewa da sanya hannu kan dacewa da sashi don sabis ✔
• B) Shirya daftarin farko na rahoton lalacewa
• C) Daidaita alamun axis na jadawali na ƙayyadaddun tsari
• D) Shirya ƙamus na ƙamus na ƙarfe don gabatarwa

Bayani: AI; Yana iya hanzarta ayyuka kamar bayanin ra'ayi, rubuta rahoto, lissafin hasashe da kwarangwal na lamba. Koyaya, amincewa da dacewa da sashe don sabis / jirgin shine yanke shawara wanda ke haifar da amincin rayuwa da dukiyoyi; ba za a iya canjawa wuri zuwa AI ba tare da ƙwaƙƙwaran gwaji da izinin injiniya mai izini ba.

2. Kun tambayi AI game da ƙarfin amfanin ƙarfe na AISI 4140 lokacin da aka kashe da zafi a 200 ° C, kuma ya amsa '1500 MPa tabbas'. Menene daidai matakin farko da za a ɗauka?

• A) Sanya darajar kai tsaye a cikin asusun ƙira saboda AI yana da tabbaci
• B) Tabbatar da ƙimar tare da takardar bayanan kayan aiki / takardar shedar niƙa da gwajin tensile idan ya cancanta, ba dogaro da ƙimar maki ɗaya ba ✔
• C) Rage darajar ba da gangan ba don zama a gefen aminci
• D) Karɓar ƙimar kuma matsawa kai tsaye zuwa girke-girke na maganin zafi

Bayani: Samfuran harshe na iya ƙirƙira lambobi marasa tabbas (hallucination). Ƙarfin ƙima; Ya bambanta akan kewayon da yawa dangane da yanayin zafin jiki, kauri na sashe da ƙarfi. Wajibi ne a sami darajar daga takardar shedar niƙa ko gwajin gwagwarmaya daidai da ma'auni, kuma kada a haɗa da fitowar AI guda ɗaya a cikin ƙira.

3. Tare da samfurin ilmantarwa na na'ura yana tsinkayar dukiyar karfe, AI ya samar da darajar 'lafiya' don haɗin gwal a waje da kewayon abun ciki a cikin bayanan horo. Menene babbar matsalar wannan fitarwa?

• A) Samfurin koyaushe yana ba da sakamako a cikin kashi, MPa ba daidai ba ne
• B) Hasashen ba shi da tabbas saboda an yi shi a waje da kewayon bayanan horo (extrapolation) kuma dole ne a tabbatar da shi ta hanyar gwaji ✔
• C) Model kawai yana aiki don bakin karfe
• D) Fitowar samfurin koyaushe yana daidai da ainihin ƙimar sau biyu

Bayani: Samfuran koyon injin suna aiki da dacewa a cikin kewayon bayanan horo (interpolation); Lokacin da ke waje da wannan kewayon (extrapolation), tsinkaya ta zama abin dogaro kuma samfurin sau da yawa baya nuna wannan. Ba za a iya amfani da ƙimar da aka samar a wani wuri a waje da rarraba bayanan horo ba tare da ingantaccen gwaji ba.

4. Kun yi amfani da AI don ƙididdige rabon lokaci a cikin allo mai nau'i biyu. Wane shigarwa yakamata AI yayi amfani dashi daidai lokacin da ake amfani da 'ka'idar lever'?

• A) Zazzabi kawai; haɗuwa ba dole ba ne
• B) Sai kawai da yawa na gami
• C) Jimlar abun da ke ciki na gami tare da ƙayyadaddun iyaka na lokaci a yanayin da ya dace ✔
• D) Yanayin narkewa kawai

Bayani: Ƙa'idar lever tana ba da adadin lokaci ta amfani da nisa tsakanin abubuwan iyakoki na lokaci (misali ƙayyadaddun lokaci mai ƙarfi da ruwa) da jimillar abun da ke ciki na gami a wani zafin da aka ba. Idan an karanta haɗin kan iyakokin lokaci ba daidai ba, rabon zai kasance gaba ɗaya kuskure; Saboda haka, tabbatar da dabi'u daga zane-zane na lokaci ya zama dole.

5. Kun auna girman hatsi daga microstructure tare da kwararar sarrafa hoto mai ƙarfin AI, amma kun manta da daidaita ma'aunin ma'auni zuwa pixel. Me yasa sakamakon bai kasance abin dogaro ba?

• A) Duk ma'aunin ma'auni suna kan sikelin da bai dace ba saboda ba a yi tsayin pixel ba ✔
• B) Ana gurbata ma'auni saboda ana amfani da hoton launi
• C) Girman hatsi za a iya auna shi kawai ta hanyar SEM, microscope na gani ba daidai ba ne
• D) Auna ba lallai ba ne saboda ƙudurin hoton yana da girma

Bayani: A cikin ƙayyadaddun ƙayyadaddun ƙayyadaddun ƙididdiga, kowane pixel yayi daidai da tsayi; Idan ba a daidaita wannan canjin tare da ma'aunin ma'auni ba, duk ma'auni kamar girman hatsi, rabon lokaci, da porosity zai bayyana akan sikelin da bai dace ba. Calibration shine mataki na farko kuma na wajibi na auna tushen hoto.

6. AI ya ba da shawarar girke-girke na maganin zafi (austenitizing + quenching + tempering) don cimma taurin manufa don karfe. Menene mafi kyawun mataki kafin amfani da takardar sayan magani?

• A) Aiwatar da girke-girke kai tsaye zuwa samar da taro
• B) Aiwatar da girke-girke akan yanki na gwaji, auna taurin / microstructure kuma kwatanta shi da ma'auni ✔
• C) Kawai ƙara daɗaɗɗen zafin jiki a gefen amintaccen
• D) Gyara da amincewa kawai rubutun rubutun

Bayani: Sakamakon maganin zafi; Kaurin sashe na sashe, ƙarfin ƙarfi (Jominy) ya bambanta da ƙarfi tare da gaskiyar tanderu da yanayin sanyaya. Rubutun da AI ta bayar shine hasashe na farko; Ya kamata a tabbatar da shi ta hanyar yin amfani da shi a kan gwajin gwaji da auna ma'auni da ƙananan tsari da kwatanta shi da ma'auni.

7. Kun tambayi AI don 'ƙayyadadden lambar abu' na ma'aunin ASTM don ƙimar ƙayyadaddun abu kuma ya amsa da ainihin ƙima. Menene daidaitaccen halayen injiniya?

• A) Rubuta ƙimar kai tsaye ga rahoton saboda AI yana da zamani
• B) Canza ƙima ba da gangan ba don zama a gefen aminci
• C) Tabbatar da lambar abu, sigar da ƙima daga daidaitaccen rubutu na hukuma da na yanzu ✔
• D) Ƙara lambar abu zuwa fayil ɗin dacewa ba tare da duba shi ba

Bayani: Ko da ƙirar harshe sun san daidaitaccen sunan daidai, ƙila su ƙila su ƙila su ƙila su daidaita lambar labarin, sigar, da wurin warwarewa. A cikin filayen ɗauri kamar ASTM/ISO/EN, kowane abu da ƙimar dole ne a tabbatar da su ta zahiri daga tushen hukuma da na yanzu; Fitowar AI kawai yana nuna inda za a duba.

8. Kun nuna hoton fashewar abin rufe fuska ga AI kuma kun sami amsar 'wannan tabbas karaya ce ta gajiya'. Menene madaidaicin hanya don bincika tushen dalilin?

• A) Rubuta ganewar asali kai tsaye a cikin rahoton saboda AI yayi magana daidai
• B) Kawai ƙara ƙuduri na hoto kuma tabbatar da ganewar asali
• C) Shigar da sabon sashi ba tare da bincikar fashe ba
• D) Tabbatar da ganewar asali a matsayin hasashe tare da fractography, tarihin caji, da shaidar microstructure ✔

Bayani: Fassarar Fassara Fassara; Fractography (striation, rairayin bakin teku, dimple/ cleavage rabuwa tare da SEM) yana buƙatar la'akari da tarihin kaya, yanayi da ƙananan shaida tare. Mahimmin ganewar asali AI daga hoto ɗaya shine hasashe; Ba za a iya bayyana tushen dalilin ba tare da tabbatarwa ta hanyar binciken dakin gwaje-gwaje ba.

9. AI ya ba da kololuwa zuwa takamaiman lokaci a cikin tsarin XRD (X-ray diffraction). Menene hanya mafi dacewa don tabbatar da gano lokaci abin dogaro?

• A) Daidaita kololuwa zuwa bayanan tsarin ƙididdiga (misali katunan PDF) da bincika zoba da kayan tarihi ✔
• B) Karɓar aikin kai tsaye saboda AI ya faɗi sunan lokaci
• C) Yin la'akari kawai mafi girman kololuwa da watsi da sauran
• D) Launi samfurin kuma sanya shi kyakkyawa na gani

Bayani: Ana yin ganewar lokaci na XRD ta hanyar daidaita ma'auni mafi girman matsayi da ƙarfi tare da sanannun alamun tunani (misali katunan ICDD/PDF). Sanya kololuwar kololuwa guda a ware yaudara ce; Ya kamata a yi la'akari da kololuwar kololuwa, fifikon fifiko da kayan tarihi kuma an tabbatar da aikin tare da bayanan bayanai.

10. Daftarin rahoton lalacewa / dubawa wanda AI ya shirya ya ƙunshi ƙimar fasaha waɗanda ke da ruwa amma asalinsu ba a sani ba. Menene daidai hali?

• A) Gabatar da rahoto kamar yadda yake saboda rubutun yana da kyau
• B) Zagaye da haskaka dabi'u
• C) Kawai karanta taƙaitawar zartarwa kuma ku tsallake sauran
• D) Daidaita kowace ƙima tare da ma'aunin asali / tushe kuma cire ✔ maras tabbas

Bayani: Kowane ƙimar fasaha a cikin rahoton; Dole ne a iya gano shi don aunawa, rikodin na'urar ƙira, daidaitattun bayanai ko bayanan masana'anta. Ko da ƙima ce a cikin jimla mai kyau, babu lambar da ba a tabbatar da tushen sa ba da za a saka shi a cikin rahoton ko tushen yanke shawara; In ba haka ba, ana canza ƙima ta ƙima zuwa takaddar da ke haifar da alhaki.

11. Menene mafi kyawun mataki kafin amfani da lissafin Python wanda aka samar da AI (misali, dangantakar Hall-Petch don girman girman hatsi)?

• A) Aiwatar da lambar kai tsaye zuwa ainihin bayanan samarwa
• B) Kawai tabbatar da lambar tayi gajere da kyau
• C) Gudanar da lambar a cikin yanayin gwaji tare da sanannun sakamako kuma tabbatar da raka'a da fitarwa ✔
• D) Share sashin lambar don rage shi

Bayani: Lambar na iya bayyana daidai amma tana ba da sakamako mara kyau: ƙila a sami kuskuren naúrar, ƙididdigewa mara daidai ko kuskuren dabara. Gudun lambar a cikin ƙaramin gwajin gwaji inda kuka san sakamakon da hannu kuma kwatanta shi da abin da ake tsammani da kuma duba raka'a yana kama kuskuren kafin ya kai ga ainihin yanke shawara.

12. Kana so ka manna your kamfanin ta ba tukuna jadadda mallaka gami girke-girke da kuma aiwatar sigogi a cikin wani jama'a samuwa AI kayan aiki da bukatar bincike. Wanne hanya ce mafi daidai?

• A) Manna girke-girke kamar yadda ake yin sauri
• B) Ya isa kawai share sunan kamfanin kuma raba duk sauran sigogi.
• C) Raba bayanan sannan kuma neman AI don share su
• D) Bincika manufofin cibiyoyi da ɓoye bayanan ko amfani da amintaccen kayan aiki ✔

Bayani: Garin girke-girke, sigogin tsari da ƙayyadaddun abokin ciniki sirrin ciniki ne da kuma mallakin hankali. Kafin raba bayanai, ya zama dole a bincika manufofin kamfani, ɓoye sunansa, ko amfani da kayan aiki na cibiyoyi/kan-site wanda baya aika bayanan.

13. Tsarin SPC mai amfani da AI (ƙididdiga tsarin ƙididdiga) ya nuna cewa an ƙetare iyakar ginshiƙi mai sarrafawa a cikin tsari na samarwa. Menene daidaitaccen halayen injiniya?

• A) Yi watsi da gargaɗin kuma ci gaba da samarwa
• B) Fadada iyakokin sarrafawa ta yadda babu gargadi
• C) Yi la'akari da gargaɗin azaman faɗakarwa, tabbatar da ma'auni, bincika tushen dalilin kuma yanke shawara tare da ma'aunin karɓa ✔
• D) Kai tsaye zazzage duka batch ɗin ba tare da aunawa ba

Bayani: Wucewa iyakar ginshiƙi mai kulawa gargadi ne; Tushen sanadin baya bayyana ta atomatik. Yana iya zama kuskuren aunawa, batun samfur, ko ainihin tafiyar da tsari. Hanyar da ta dace ita ce la'akari da gargaɗin a matsayin mai faɗakarwa, tabbatar da ma'auni, bincika tushen dalilin kuma yanke shawara bisa ka'idojin karɓa da sakamakon NDT / jarrabawa.

14. Ƙimar hoto ta AI-taimaka akan ɓangaren simintin gyare-gyaren da aka yiwa alama yanki a matsayin 'm'. Menene madaidaicin hanya don sashi mai mahimmancin aminci?

• A) Yin la'akari da kimantawa na farko na AI azaman kayan aikin nunawa da haɗa yarda ta ƙarshe zuwa sakamakon NDT da izini mai izini ✔
• B) Yi la'akari da kimantawar AI a matsayin karɓa na ƙarshe kuma yarda da ɓangaren
• C) Ci gaba zuwa tsari na gaba ba tare da bincika sashin ba kwata-kwata
• D) Ya isa kawai duba saman sashin.

Bayani: Simintin gyare-gyare na ciki (porosity, shrinkage, inclusions) sau da yawa ba a iya gani daga saman kuma ana kimanta su ta hanyar gwaji mara lalacewa (NDT: radiyo, ultrasonic). Ƙimar farko da AI ke bayarwa ta hanyar hoton shine kayan aikin nunawa; Karɓar ƙarshe na ɓangaren mahimmin aminci ana bayar da shi tare da NDT da izini na dubawa daidai da ma'auni.