پديد آورنده :
شهرياري، مرضيه
عنوان :
مدل سازي خصوصيات آكوستيك پارچه هاي پرده اي پلي استري با استفاده از هوش مصنوعي
مقطع تحصيلي :
كارشناسي ارشد
محل تحصيل :
اصفهان: دانشگاه صنعتي اصفهان، دانشكده نساجي
صفحه شمار :
ده،97ص.: مصور،جدول،نمودار
يادداشت :
ص.ع.به فارسي و انگليسي
استاد راهنما :
محمد ذره بيني
توصيفگر ها :
جاذب متخلخل , ضريب جذب صوت , لوله امپدانس , شبكه عصبي مصنوعي
تاريخ نمايه سازي :
2/7/92
استاد داور :
حسين حسني، داريوش سمناني
چكيده فارسي :
به فارسي و انگليسي: قابل رويت در نسخه ديجيتالي
چكيده انگليسي :
Isfahan University Of Technology Department Of Textile Engineering Modeling Of Acoustic Properties Of Polyester Curtain Fabrics Using Artificial Intelligant Marzieh Shahriari ma shahriari@yahoo com Department of Textile Engineering Isfahan University of Technology Isfahan 84156 83111 Iran Degree M Sc Language Farsi Supervisor s Name and Email address Mohammad Zarrebini zarrebini@cc iut ac ir Abstract Despite all modern technological achievements creation of acoustically acceptable environment both at work and resident has become a challenging issue for human being Elimination and reduction of unwanted noise can be achieved by vast array of porous materials one of which are textile fabrics Textiles in addition to conventional end use of apparels are used as curtains floor coverings and items of upholstery In these applications textiles can serve as sound absorbent material Currently non woven and knitted fabrics are used as sound absorbent in various end uses In this work acoustic properties of woven polyester fabrics were investigated at various frequencies using impedance tube apparatus Additionally in order to study the effect of fabric layering on sound absorbency experiments using three layered fabric systems were conducted Results were analyzed using commonly used statistical methods Results showed that factors such as warp fineness and warp fabric density together with factors such as weft yarn twist and type in addition to fabric cover factor influence Noise Absorption Coefficient NAC at 2000Hz It was found that while weft yarn type and twist affect NAC at 250Hz NAC is affected by fabric mass at 1000Hz Results also confirmed that both layering of fabric and used frequency affect fabric NAC Parameters such as fabric thickness cover factor number of layers and frequency were selected as inputs to the multi layer feed forward Error Back Propagation algorithm used for modeling and predicting of NAC of the samples In order to achieve the best prediction 17 structures were used and on the basis of Cross validation technique 270 data in 8 groups were classified Finally a network with 2 hidden layers 19 units in the first layer and 10 units in the second layer was chosen by error and trial as the best model The model was capable of providing NAC with the least average Mean Square Error MSE for both train and test data In this model Correlation Coefficient R value between predicted and measured values of NAC and MSE of test data was found to be 0 84 and 0 0028 respectively Efficiency of model was explored by choosing 20 data randomly and calculating percent of error between measured and predicted values of NAC Average maximum and minimum of percent of errors were found to be 5 50 16 66 and 0 00 respectively The results confirmed the high reliability of the model in prediction of noise absorption coefficient of polyester curtain fabrics Key Words Acoustic porous absorbent noise absorption coefficient impedance tube neural network polyester curtain fabrics PDF created with pdfFactory trial version www pdffactory com
استاد راهنما :
محمد ذره بيني
استاد داور :
حسين حسني، داريوش سمناني
