应用人工神经网络分析织物的形态与性能之间的关系  

Analysis of the shape-property relationship of fabric by using artificial neural network

作　　者：余志才 钟跃崎[1,2] YU Zhicai;ZHONG Yueqi(College of Textiles,Donghua University,Shanghai 201620,China;Key Laboratory of Textile Science&Technology,Ministry of Education,Donghua University,Shanghai 201620,China)

机构地区：[1]东华大学纺织学院,上海201620 [2]东华大学纺织面料技术教育部重点实验室,上海201620

出　　处：《毛纺科技》2021年第5期1-7,共7页Wool Textile Journal

基　　金：国家自然科学基金项目(61572124)。

摘　　要：为了探索织物的三维悬垂形态与织物力学性能之间的关系,首先利用BP(Back Propagation)神经网络研究织物三维悬垂形态与织物弯曲刚度、剪切刚度之间的相互映射机制;然后探讨了利用分类模型直接判定三维悬垂模型对应的织物柔软程度的可行性。结果表明:根据织物的弯曲刚度、剪切刚度和面密度,对织物的悬垂性能可进行较高精度的预测,在已知织物面密度和三维悬垂形态的前提下,也可通过形态特征预测织物弯曲刚度和剪切刚度。在此基础上,利用分类模型可以在仅输入悬垂形态的情况下判定织物的柔软程度,其准确率相当于人类大脑识别率的97.8%。In order to explore the relationship between the three-dimensional drape(shape)and the mechanical properties of woven fabric,the mapping mechanism between three-dimensional drape and the shear stiffness/bending rigidity of fabric was studied by using a BP(Back Propagation)neural network.Moreover,the feasibility of using the classification model to directly predict the softness of the fabric corresponding to the three-dimensional drape model was discussed.The results show that it is possible to predict the fabric drape based on the bending rigidity,shear stiffness and fabric weight.The bending rigidity and shear stiffness of the fabric can also be predicted according to the fabric weight and its 3D drape.Additionally,the classification models can be used to classify fabric softness when only the drape form is input.The prediction accuracy was equivalent to 97.8%of human brain recognition.

关 键 词：人工神经网络 织物悬垂形态 力学性能 织物柔软程度 

分 类 号：TS104.7[轻工技术与工程—纺织工程]