Research And Application Of Intelligent Method For Garment Pattern Generation

Apparel CAD system is an important link in the information management of apparel production enterprises.The use of apparel CAD board can improve the production efficiency of enterprises and respond to market changes quickly.However,the existing clothing CAD system is lack of intelligence,which can be regarded as converting traditional manual drawing template tool into a mouse and keyboard,the template drawing process has no substantial change,still need the experience of the template division.Therefore,in order to better adapt to the needs of enterprises,it is necessary to build intelligent clothing CAD.This paper proposes a method for intelligently generating clothing patterns,which is mainly for the conventional production methods of clothing enterprises.The intelligence of the method is mainly embodied in two aspects:on one hand,it has the function of automatic board-playing,which can realize the automatic generation of the template of a certain type of clothing,and can also realize the corresponding change of the pattern of the clothing model for the clothing of the changed style;On the other hand,it has the habit of self-study.It proposed a method to quantify the experience of pattern maker to extract the experience of pattern maker,and then used BP neural network to train for the learning model,so as to learn the drawing experience of experienced pattern maker and apply it to the automatic generation process of garment pattern.Based on the above methods,a method for generating the intelligent sample of conventional men's shirts is constructed.The research is mainly carried out from the following aspects:(1)By collecting and analyzing the styles of conventional men's shirts,dividing their parts,and obtaining the basic styles and variation styles;(2)Select a suitable template drawing method for template drawing;(3)Conduct a comparative analysis of style changes and corresponding structural changes,and extract the key control points of sample changes;(4)This paper studies the change rule of the structure of conventional men's shirt when its style changes,and focuses on the experimental study of the change rule of three key parts,including the change of armhole curve,the change of sleeve length with the change of cuff and elbow volume,and the relationship between armhole and sleeve hill curve.(5)A quantitative method for drawing master's drawing experience is proposed,and a self-learning model based on BP neural network is constructed.Finally,based on the research on the intelligent generation method of the conventional men's shirt model,the object-oriented method is used to digitally describe its style and structure,thus constructing the conventional men's shirt smart boarding system.