基于面料力学性能的半紧身裙造型特征预测  

Predicting the Semi-fitting Skirt Appearance through Fabric Mechanical Properties with Artificial Neural Networks

作　　者：唐虹[1] 张渭源[2] 

机构地区：[1]南通大学纺织服装学院,江苏南通226007 [2]东华大学服装学院,上海200051

出　　处：《南通大学学报（自然科学版）》2007年第3期47-51,共5页Journal of Nantong University(Natural Science Edition)　

基　　金：南通大学自然科学基金项目(05Z070)

摘　　要：文章在对面料力学性能分析的基础上,建立了半紧身裙造型特征预测的BP神经网络模型.提出了半紧身裙造型特征为形式感、体量感、平衡感、运动感、悬垂感、节奏感、稳重感、均匀感、贴体感、平服感等.FAST系统测量的面料力学性能指标作为神经网络模型的输入层,18位专家组成的评判组对41种不同面料构成的半紧身裙造型特征作出的评价值作为BP神经网络的输出层.通过对任选8只样本的预测值和实测值的线性回归分析发现,该模型有较好的预测精度,其中贴体感、均匀感、运动感的预测精度最好;形式感、体量感、悬垂感、平服感的预测精度次之;平衡感、节奏感、稳重感的预测精度较低.The appearance of semi-fitting skirt is very important for its design and consumption. The parameters featured as shape, volume, balance, movement, drape, rhythm, firmness, uniformity, body-fitting and smoothness can be assessed by apparel experts. The prediction model of the appearance of semi-fitting skirt was built up with BP neural network. Eleven mechanical properties carried by FAST system are the input variable candidates; the output layer is ten subjective scales got from eighteen apparel experts. It is proved the network model has an excellent precision and it is an efficiency way to predict the appearance of semi-fitting skirt.

关 键 词：半紧身裙 造型特征 面料力学性能 神经网络模型 

分 类 号：TS941.717.8[轻工技术与工程—服装设计与工程]