Prediction Model of Sewing Technical Condition by Grey Neural Network.pdfVIP

维普资讯 JoumalofDonghuaUniversity (Eng．Ed，)Vo1．24．No．4(2007) 565 Prediction M odel of Sewing Technical Condition by Grey Neural Network DONGYing(董英) ，FANGFang(方方)，ZHANGWei—yuan(张渭源) FashionInstitute，DonghuaUniversity，Shahgkai 20005]，(3Jina Abstract：Thegreysystem theoryand theartificialneural technicalconditions，needleand stitch，based on fabrics’ network technology were applied to predict the sewing mechanicalproperties．Itprovidesthetheoreticreasonfor tcehn icalcondition．Therepresentativeparameters，suchas thetraditionalvalue． nedele， stitch， were selectde ． Prediction model was Artificialneuralnetwork technology succeedsin the established based on the different fabrics’ mechanical textileandgarmentfieldsbecauseofitsexcellentcapability propertiesthatmeasuredbyKES instrument．Greyrelevant of dealing with non—linear complicated problems． Grey de~eeanalysiswasapplide tochoosetheinputparameters system theory fitsfortheproblemsthathaveincomplete ofthe neuralnetwork．The resultshowed thatprdeiction information． W e created the grey neural network mod elhas good precision．Theaverage relative errorwas prediction modelby combining these two theories．Grey 4．08％ f0rneedleand4．25％ f0rstitch． relevantdegree analysiswas applied to select the input Keywords：grey relevantdegree；neuralnet~grk；needle； parametersofartificialneuralnetwork． stitch：KESmeasure~Tlenl：predictionmodel CI number：TS941．61 Documentcode：A 1 ~ rovoqtofFabrics’[Vk~m imlP旧3erti ArticleID：1672—5220(2007)04—0565—04 Forty—five kinds of woven fabricswere collected，