主成分分析与BP神经网络对束裤压力的预测

详细信息    查看全文 | 推荐本文 |

英文篇名：Principal component analysis and BP neural network prediction of beam pants pressure 作者：郭颖 ; 丁正生 ; 马秋瑞 ; 林强强 英文作者：GUO Ying;DING Zhengsheng;MA Qiurui;LIN Qiangqiang;College of Sciences,Xi'an University of Science and Technology;College of Fashion and Art Design,Xi'an Polytechnic University;College of Mechanical and Electrical Engineering,Xi'an Polytechnic University; 关键词：束裤压力 ; 主成分分析 ; BP神经网络 ; BP神经网络预测 英文关键词：pressure values;;principal component analysis;;BP neural network;;BP neural network prediction 中文刊名：MFKJ 英文刊名：Wool Textile Journal 机构：西安科技大学理学院;西安工程大学服装艺术设计学院;西安工程大学机电工程学院; 出版日期：2019-07-16 出版单位：毛纺科技 年：2019 期：v.47;No.373 基金：国家自然科学基金(71473194) 语种：中文; 页：MFKJ201907013 页数：5 CN：07 ISSN：11-2386/TS 分类号：61-65

摘要

为了预测人体穿上束裤时身体各部位的压力变化,提出了一种主成分分析与BP神经网络相结合的预测方法。首先测量了90位女大学生在穿着同一品牌、款式束裤时,身体各部位所对应的压力值,然后用主成分分析的方法对影响压力的9个因素进行提取分析,得到体重、臀围和大腿中部围这3项作为神经网络的输入指标,压力值作为输出指标。借助Matlab软件自行编程并进行调试,比较并分析BP神经网络与PCA-BP神经网络的预测结果。结果显示,PCA-BP神经网络较BP神经网络预测的相对误差的精度提高了3. 790 7%,相对误差的绝对值精度提高了5. 793 4%,训练时间减少了18. 321 s。
        In order to predict the pressure change of the body when wearing the trousers,the prediction method combining principal component analysis and BP neural network was proposed. First,the pressure value corresponding to a female college student when wearing a pair of trousers in the same brand was measured. Then,nine factors affecting the pressure were extracted and analyzed using the principal component analysis method,the three indicators of body weight,hip circumference and middle thigh were obtained as the input of the neural network,and the pressure value was used as the output.Using self-programming program to debug with matlab,the prediction results of general BP neural network and PCA-BP neural network was compared and analyzed. The results show that the accuracy of the relative error of the PCA-BP neural network was improved compared to the general BP neural network,the accuracy of the absolute value of the relative error was improved,and the training time was reduced by seconds.

引文

[1]张薇薇,李艳梅.服装压力舒适性研究进展[J].上海工程技术大学学报,2012,26(1):46-49.
    [2]罗笑南,曾龙,聂卉.三维紧身内衣的压力分布计算模型[J].中山大学学报(自然科学版),2001,40(6):99-101.
    [3]FAN J,CHAN A P.Prection of gridle's pressure on human body from the pressure measurement on a dummy[J].International Journal of Clothing Science&Technology,2005,17(1):6-12.
    [4]由芳,张欣.紧身服的宽裕量及弹性模量与服装压感的关系[J].西安工程科学学院学报,2000,14(2):133-137.
    [5]WANG J M,LUO X N,LI Y,et al.The application of the volumetric subdivision scheme in the simulation of elastic human body deformation and garment pressure[J].Textile Research Journal,2005,75(8):591-597.
    [6]ZHANG X,YEUNG K W,LI Y.Numerical simulation of 3D dynamic garment pressure[J].Textile Research Journal,2002,72(3):245-252.
    [7]MCGHEE D E,STEELE J R.Breast elevation and compression decrease exercise-induced breast discomfort.[J].Med Sci Sports Exerc,2010,42(7):1333-1338.
    [8]张缓缓,赵娟,李仁忠,等.k-means聚类算法在织物疵点检测中的应用[J].毛纺科技,2016,44(3):11-14.
    [9]李鹏飞,陈耀武,景军锋,等.基于DSP的实时织物疵点检测系统[J].毛纺科技,2015,43(10):47-52.
    [10]闵惜琳,刘国华.用MATLAB神经网络工具箱开发BP网络应用[J].计算机应用,2001,21(s1):163-164.
    [11]中国肥胖问题工作组,季成叶.中国学龄儿童青少年超重、肥胖筛查体重指数值分类标准[J].中华流行病学杂志.2004(2):97-102.
    [12]王学仁,王松桂.实用多元统计分析[M].上海:上海科学技术出版社,1990:270-344.
    [13]张心心.MATLAB软件在信号与系统实践教学中的意义[J].通讯世界,2016(12):61.
    [14]常青霞,高卫东.基于BP神经网络的国际服装流行色短期预测[J].毛纺科技,2018,46(2):87-91.
    [15]沈波.基于BP神经网络的股价预测应用研究[D].长沙:湖南大学,2010.
    [16]孙鸿文,陈庆官.BP神经网络识别服装面料成形性探讨[J].苏州大学学报(工科版),2005,25(1):60-63.
    [17]沈加加,周翔,屠天民.基于BP神经网络的色纺纱配色[J].纺织学报,2015,36(11):34-38.
    [18]刘为敏,谢红.BP神经网络下的智能化合体服装样板生成[J].纺织学报,2018,39(7):116-121.
    [19]王海峰,姜雲腾,李萍.考虑气象因素的PCA-BP神经网络短期负荷预测[J].电工电气,2018(7):38-41.