A wani gagarumin ci gaba a fannin binciken injiniya, wani sabon bincike ya nuna ingancin haɗa siginar modulation bispectrum (MSB) da hanyoyin sadarwa na jijiyoyi masu motsi (CNN) don gano lahanigiyar bevel mai karkaceWannan sabuwar hanyar ta yi alƙawarin inganta daidaito, ganowa cikin sauri, da kuma tsarin bincike mai wayo don manyan akwatunan gearbox da ake amfani da su a cikinaikace-aikacen sararin samaniya, motoci, da masana'antu.

Karkacegiyar bevelmuhimman abubuwan watsawa ne da ake samu a cikin injunan aiki masu ƙarfi, jiragen sama masu saukar ungulu, tsarin turawa na ruwa, da kuma na'urorin rage ƙarfin masana'antu. Saboda yanayinsu mai rikitarwa da yanayin aiki, gano lahani na gear da wuri kamar rami, lalacewa, da karyewar haƙori ya kasance ƙalubalen fasaha. Dabaru na sarrafa sigina na gargajiya galibi suna fama da tsangwama da halayen lahani marasa layi.

Sabuwar hanyar ta gabatar da tsarin gano kurakurai a matakai biyu. Da farko ana nazarin siginar girgiza da tsarin kayan aiki ke samarwa ta amfani da bispectrum na siginar modulation (MSB) wata dabarar nazarin spectral mafi girma wacce ke kama fasalulluka marasa layi da marasa Gaussian na siginar yadda ya kamata. MSB tana taimakawa wajen bayyana halayen kurakurai masu sauƙi waɗanda aka daidaita waɗanda galibi ake ɓoye su a cikin mitoci na yau da kullun.

Na gaba, ana canza bayanan siginar da aka sarrafa zuwa hotunan mitar lokaci kuma ana tura su zuwa hanyar sadarwa ta jijiyoyi ta convolutional (CNN) wani tsari mai zurfi na koyo wanda zai iya cire fasalulluka na kurakurai masu girma ta atomatik da kuma rarraba yanayin kayan aiki. An horar da wannan samfurin CNN don bambance tsakanin kayan aiki masu lafiya, ƙananan kurakurai, da mummunan lalacewa a cikin yanayi daban-daban na kaya da gudu.

Sakamakon gwajin, wanda aka gudanar a kan wani na'urar gwajin gear mai siffar karkace ta musamman, ya nuna cewa hanyar MSB CNN ta cimma daidaiton rarrabuwa sama da kashi 97%, ta fi hanyoyin gargajiya kamar nazarin FFT da ma wasu dabarun koyo masu zurfi waɗanda suka dogara da bayanan girgizar ƙasa. Bugu da ƙari, wannan samfurin haɗin gwiwa yana nuna ƙarfi mai ƙarfi ga hayaniyar bango, wanda hakan ya sa ya dace da aikace-aikacen masana'antu na gaske.

Haɗa tsarin siginar daidaitawa tare da CNN ba wai kawai yana haɓaka aikin gane lahani ba, har ma yana rage dogaro da injiniyan fasali na hannu a al'ada, wanda ke ɗaukar lokaci da ƙwarewa dangane da tsari. Hanyar tana da ƙwanƙwasa kuma ana iya amfani da ita ga sauran kayan aikin juyawa, kamar bearings dagiyar taurari.

Wannan bincike yana wakiltar ci gaba a cikin haɓaka tsarin gano kurakurai masu hankali don Masana'antu 4.0 da kuma faffadan fannin kera kayayyaki masu wayo. Yayin da sarrafa kansa da amincin injina ke ƙara zama mahimmanci,