Research On The Auto-generation Technology Of Children’s Clothing Paper Pattern Based On BP Neural Network

The apparel industry is developing rapidly and digitalization has become the future direction of the apparel industry.Traditional garment CAD technology has a long production cycle and a lot of repetitive labor,requires a high level of experience for garment pattern makers,and cannot produce garment paper samples personalized for each consumer.Children are a group with a unique personality for clothing needs,in the growth period,and their height and size will change greatly every year.With the improvement of consumers’ living standards,the demand for personalized clothing for children will increase year by year.How to save labor,improve the efficiency of children’s clothing plate making,meet the personalized needs of children’s clothing and improve the fit of children’s clothing paper samples is an urgent problem for clothing enterprises.In this paper,we simulate the patternmaking experience of children’s clothing pattern designers through BP neural network model,edit the program of children’s clothing pattern automatic generation with Visual-LISP language embedded in Auto-CAD,and input the predicted children’s clothing pattern detail size results into the DCL dialog box developed by Auto-CAD twice.Firstly,351 net body size data of children aged 3-6 years old measured manually with Martin’s measuring tape and soft tape were obtained from the database system,and the SPSS software was used to analyze the basic statistics,correlation analysis and regression analysis to obtain the distribution pattern of net body size data of children,the correlation between key parts of children and body size of each part,and establish the linear regression equation between key parts of children and other The linear regression equation of children’s key parts and other parts was established.The paper takes Donghua child basic block 2020 edition,children’s basic skirt type and children’s jeans as examples respectively,and builds children’s clothing paper pattern size prediction model by simulating the design experience of clothing paper pattern maker through BP neural network algorithm based on a large amount of children’s net body size data,and then builds children’s clothing paper pattern parametric model according to the clothing pattern making rules.On the Auto-CAD2020 platform,the embedded Visual-LISP programming language is used to write the automatic drawing program for the parametric children’s clothing paper pattern,and the size data of children’s clothing paper pattern predicted by BP neural network is input into the DCL dialog box to realize the automatic generation of children’s clothing paper pattern,and the operator only needs to obtain the net body size parameters of key parts of human body to quickly generate children’s clothing paper pattern.The paper evaluates the effect of the generated paper samples of Donghua child basic block 2020 edition,children’s clothing basic skirt type and children’s jeans through CLO3 D virtual fitting interface display and real person trying on after the garment is made into a garment to verify the rationality of the automatic generation system.The results show that the automatic generation system of clothing paper samples can get rid of the reliance on the experience of clothing pattern makers,shorten the production cycle of clothing enterprises,realize personalized mass customization of clothing enterprises,and provide experience and theory for the intelligent production of clothing enterprises.