ارائه ابزار محاسباتی نرم مبتنی بر مدل‌های ترکیبی به‌منظور بهبود پیش‌بینی کیفیت منسوجات تولیدی در صنعت پوشاک

نوع مقاله : مقاله پژوهشی

نویسندگان

1 Department of Industrial Engineering, Isfah an University of Technology, Isfahan, Iran

2 دانشکده مهندسی صنایع و سیستم ها، دانشگاه صنعتی اصفهان، اصفهان، ایران

چکیده

ب شبکه‌های عصبی مصنوعی ابزارهای پیش‌بینی دقیقی برای دامنه وسیعی از مسائل هستند که نیاز به داده‌های زیاد برای حصول نتایج دقیق، کاربرد آن‌ها را با محدودیت مواجه کرده است. این درحالی است که فراهم آوردن داده‌های مورد نیاز به منظور ارائه پیش‌بینی‌های دقیق با شبکه عصبی مصنوعی در صنعت نساجی، اصولاً بسیار هزینه‌بر و زمان‌بر است. از این رو، استفاده از روش‌هایی که قادر به ارائه پیش‌بینی با تعداد داده‌های قابل حصول کم هستند، در این‌گونه از صنایع مناسب‌تر‌ و کارآمدتر خواهد بود. در این مقاله، از ترکیب روش‌های شبکه‌های عصبی مصنوعی و رگرسیون فازی به ارائه یک مدل هوشمند نرم به منظور پیش‌بینی کیفیت درز پوشاک تولیدی پرداخته شده است. ایده اصلی روش پیشنهادی استفاده از مزایای محاسبات نرم مجموعه‌های فازی به‌منظور حصول نشخه‌ای بهبودیافته از شبکه‌های عصبی مصنوعی در شرایط داده‌های قابل حصول کم است. نتایج بدست آمده از بکارگیری روش پیشنهادی در پیش‌بینی کیفیت درز پوشاک، بیانگر عملکرد بالاتر این روش درتقابل با مدل‌های تشکیل‌دهنده خود و همچنین سایر روش‌های ترکیبی موجود است.

کلیدواژه‌ها


عنوان مقاله [English]

Proposing a soft hybrid computing model to improve the prediction of the quality of seams in garments

نویسندگان [English]

  • Mehdi Khashei 1
  • Nesa Ahmadyar 2
1 Department of Industrial Engineering, Isfah an University of Technology, Isfahan, Iran
2 Department of Industrial and Systems Engineering, Isfahan University of Technology, Isfahan, Iran
چکیده [English]

For training the artifcial neural network to accurately predict a specifc parameter, several experimental results ar
generally needed which makes it costly and time-consuming. Hence, there is a need for developing other method
that can accurately predict based on a small number of experimental data. In this paper, a combination of artifci
neural network and fuzzy regression methods was employed to develop a soft intelligent model for predicting th
quality of seams in garments. The main idea of the proposed method is to simultaneously use the advantages of so
computing of fuzzy sets to achieve improved results from artifcial neural networks based on a relatively small ex
perimental dataset. The results obtained from the proposed model showed its higher performance in comparison t
its constituent models as well as other existing combinational methods.






کلیدواژه‌ها [English]

  • Multilayer Perceptron Neural Networks
  • Fuzzy regression
  • Hybrid models
  • Textile industry
  • Predict quality of clothing seam
 ,Thesis Sc .M ,"fabrics suiting of Sewability ".,K ,Choudhry 1.
1995. ,Delhi of University
 ,Seams and Stitches for Practice Standard 6193-16,D ASTM 2.
.www ,PA ,Conshohocken WEst ,Internathional ASTM
2016. ,org.astm
.3 م، شیخ زاده،. م، حجازی. و ر، شکریان. تحلیل ویسکواالستیک پارچه های درزدار
تحت بارگذاری ثابت نســبت به زمان. دولتی - وزارت علوم، تحقیقات، و فناوری -
دانشگاه صنعتی اصفهان - مهندسی نساجی. .1395
 to networks neural artificial of Application ".,al et .,C.P ,Hui 4.
 :(5)19 .Technology and Science Clothing of Journal tionalInterna ."fabrics of performance sewing of prediction the
2017. 291-318, .pp
 on Investigation ",Petrov .H and .D ,Krasteva-Germanova 5.
 International .s"fabric light of sewing by quality s'seam the
57- .pp :)1(20 .Technology and Science Clothing of Journal
2008. 64,
 garment of evaluation objective the Towards ".,al et .,J ,Fan 6.
 and Science Clothing of Journal International ."appearance
1999. 151-160, .p :)2/3(11 ,Technology
169- .pp :)3/4(14 ,Technology and Science Clothing of nalJour International ."quality appearance garments of dictionpre qualitative for system the of Development ".,J ,Geršak 7.
2002. 180,
 garment in problems sewability of Prognosis ".G ,Stylios 8.
 Systems in .technology based-computer using manufacture
1990. .on Conference International IEEE ,"Engineering
 and indicators pucker Seam ",Dobilaitė .V and .M ,Juciėnė 9.
 ."machine sewing a of parameters the upon dependence their
 ,Technology and Science Clothing of Journal International
2008. 231-239, .pp :)4(20
 of control quality automated for images fabric from turesfea relevant Selecting ",Zeng .X and ,Miou .C.J .,L ,Koehl 10.
 ,"grading experts human and analysis data using puker seam
2007. 39-54, .pp .Springer ,textile Computational in
 control quality Automatic ",Dermatas .S.E and .G.I ,Mariolis 11.
 International ."detection shadow on based puckers seam of
 ,Technology and Engineering Aided Computer of Journal
2009. 225-238. .pp :)2(1
 Part :system expert fabric worsted A ",Hunter .L and .J ,Fan 12.
 .pp :)9(68 ,Journal Research Textile ."development System :I
1998. 680-686.
 The ",Jayaraman .S and ,Rajamanickam .R .,M ,Ramesh 13.
 neural artificial using by properties tensile yarn of prediction
459- .pp :)3(86 ,Institute Textile the of Journal ."networks
1995. 469.
 of performance seam Predicting ",Ng .F.S and .L.C ,Hui 14.
 ,journal research Textile ."networks neural artificial and sionregres logarithm multiple using fabrics woven commercial
1649-1657. .pp :)18(79 2009.
 in propagation Back ",Fuller .D.J and .G ,Lachtermacher 15.
 .pp :)4(14 ,forecasting of Journal ."forecasting series-time
1995. 381-393,
 time for model regression fuzzy and networks neural tificialar hybrid new A ",Bijari .M and ,Hejazi .R.S .,M ,Khashei 16.
769- .pp :)7(159 ,systems and sets Fuzzy ."forecasting series
 2008. 786,
 domestic gross of estimates three Combining ".,J.D ,Reid 17.
1968. 431-444, .pp :)140(35 ,Economica ."product
 :(4)20 ,Society Research Operational the of Journal ."castsfore of combination The ",Granger .W.C and .M.J ,Bates 18.
1969. 451-468, .pp
 of Prediction ",Khandual .A and ,Luximon .A .,P ,Kumar Ajit 19.
 neural artificial using fabrics woven cotton of profile drape
 Research Textile ."method regression multiple and network
2010. 559-566. .pp :)6(81 ,Journal
 color a of Development ",Park .K.C and ,Kim .S .,P.J ,Hwang 20.
 using printing textile transfer digital for algorithm matching
 Textile ."regression multiple and network neural artificial an
2015. 1076-1082, .pp :)10(85 .Journal Research
 and detecting for model intelligent An ".,al et .,H.Y ,Zhang 21.
 Research Textile ."network neural Elman the and gorithmsal genetic using defects fabric textured-color classifying
2011. 1772-1787, .pp :)17(81 .Journal
 of Application ",Sarkar .B and ,Majumdar .P .,A ,Majumdar 22.
 cotton of prediction the for system fuzzy-neuro adaptive an
 2005. 55-60, .pp :)1(96 .Institute tileTex the of Journal ."properties fibre HVI from strength yarn
 using hairiness yarn cotton of Modeling ".,A ,Majumdar 23.
2010. ."system inference fuzzy-neuro adaptive
 predict to model a :clustering subtractive with system enceinfer fuzzy-neuro Adaptive ",Apte .S and .S.L ,Admuthe 24.
 :(9)80 .Journal Research Textile ."relationship yarn and fiber
2010. 841-846, .pp
 prediction hand fabric for approach new A ".,al et .,Y ,Yu 25n feature based-algorithm neighbor nearest a with
2011. 574-584, .pp :)6(81 .Journal Research Textile ."scheme
 rigidity bending Predicting ",Guruprasad .R and .B ,Behera 26.
 .pp :)11(103 .Institute Textile The of Journal .)"ANFIS (temsys inference fuzzy-neuro adaptive using fabrics woven of
2012. 1205-1212,
 modeling for techniques Intelligent ".,al et .,G.-E.S ,Jeguirim27.
 Engineered of Journal ."fabrics knitted of measurements talinstrumen and attributes sensory between relationships the
2012. , (3)7 .(JEFF (Fibers & Fabrics
 ."system inference fuzzy-neuro adaptive modular using tionmigra fiber modeling :III Part .yarns spun-ring ventionalcon and ,compact ,solo ,siro on study A ".,al et .,P ,Soltani 28.
 2013. 755-765, .pp :)7(104 .Institute Textile The of Journal
 Journal .)"GEP (Programming Expression Gene and) FISAN (System Inference Fuzzy-Neuro Adaptive Using lingModel Strength Yarn" ,Moghassem .A and .A ,Fallahpour 29.
2013. 6-18, .pp:)4(8 .)JEFF (Fibers & Fabrics Engineered of
 in concentrations dye Estimating ".,A ,Nateri-Shams 30.
 ."technique fuzzy-neuro by wastewater component-three
 920-926, .pp :)9(105 .Institute Textile The of Journal The
2014.
 Ultraviolet Modeling ",Das .A and ,Majumdar .A .,P ,Hatua 31.
2014. 99-106, .pp :)3(9 .)JEFF (Fibers & Fabrics neeredEngi of Journal ."Approaches Computing Soft Using ricsFab Woven Blended Cotton-Polyester of Factor Protection
 Prediction The ",Etrati .M.S and ,Najar .S.S .,E ,Haghighat 32.
 Using Fabrics Denim Woven in Force Penetration Needle of
 & Fabrics Engineered of Journal ."Models Computing Soft
2014. 45-55, .pp :)4(9 .)JEFF (Fibers
 network neural artificial of Comparison ".,al et .,T ,Hussain 33.
 the predicting for system inference fuzzy-neuro adaptive and
2015. 934-938, .pp :)9(106 .Institute tileTex the of Journal The ."fabrics woven of recovery wrinkle
 for method intelligent An ",Koehl .L and ,Zeng .X .,Z ,Xue 34.
 s’men for formability fabric of prediction and evaluation the
2018. 438-452, .pp :)4(88 .Journal Research Textile ."suits
 series time for models ARIMA and networks neural ficialarti of hybridization novel A ",Bijari .M and .M ,Khashei 35.
2664- .pp :)2(11 .Computing Soft Applied ."forecasting
2011. 2675,
 and Network Neural Hybrid A ",Meng .M and .X ,Wang 36.
 .JCP ."Forcasting Consumption Energy for Model ARIMA
2012. 1184-1190, .pp :)5(7
 .R.S ,Hejazi and .M ,Bijari .,F.R ,Mokhatab .,M ,Khashei 37.
 Engineering Industrial of Journal ,"forecasting rate changeex foreign for model intelligence computational hybrid A"
2011. 15-29, .pp ,)15(7 .International
 Neural Artificial Improved Using Forecasting Rate changeEx Foreign ".M ,Bijari .F ,Rafiei Mokhatab .M, ،Khashei 38.
2011. ,html121._08IIEC08-IIECPaper/com.civilica.www://https ,Conference Engineering alIndustri International Eighth ,"Data Incomplete in Networks
 .pp 27: .Systems Based-Knowledge ."analysis data pressionex gene in problem classification the to approach telligentin fuzzy A ",Bijari .M and ,Hamadani .Z.A .,M ,Khashei 39.
2012. 465-474