A Study On Comfortable Performance Of Sport Brassieres Based Of Artificial Neural Network

With society changes and the progress of the times,the continuous improvement of people's living standard and the development of sports competition,sport brassieres occupy a more and more important position in modern women's life and consumers pay a increasingly attention to the comfort of sport brassieres. By the means of BP artificial neural network,the article established a model on the basic parameters,performance and comfort of the fabric to research the thermal-wet and pattern pressure comfort of the sport brassieres. According to sports science and ergonomics theory,with zoning design concept applied,this paper used dressing test to compare the experimental brassieres and the ones selected in the market with the same profile-type.Then studied how the comfort of sport brassieres affected by the fabric performances through a series of experiments.By analysis of the development status quo of recreational sport brassieres and searching and summarizing the related literature both at home and abroad, the paper provided an overview of the concept of dressing comfort and its requirements and the impact of factors. This topic selected 10 kinds of knitted fabric commonly used for sports bras and made use of equipments related to the experimental test of the basic parameters of the fabric, thermal-wet comfort performance and comfortable index of pressure form through experimental methods.Drawn up with the data, experimental officers input network training programs and submitted them to the network training by Matlab applications in the light of Matlab language rules. Input parameters as the material composition, weight, thickness and the longitudinal density value;Use of trial and error method to determine the hidden layers and the number of hidden layer neurons;output parameters as heat rates, ventilation rates, high wicking; largest training number was 1000000; target error was 0.01;momentum item:0.9;learning rate: 0.5; finally selected the three-BP network model of 4×9×1 to analyze and forecast the thermal-wet comfort and form pressure performance of the fabric.On the basis of the different sport sweating intensity and the degree flexibility pressure on various parts of representation on women's body during moderate-intensity sport,the paper zoning designed the representative parts on the upper body as chest, shoulder,lower chest and back-different parts using fabric of different comfort properties. Specific as follows: shoulder to choose high-elastic and moisture performance fabric medium; under the bust can choose good elasticity and flexibility materials; back to choose moderate flexibility and good wet performance fabric; chest to choose high-bomb and high-moisture outer cloth and interlining cloth with good moisture absorption and soft fabric. Structure-the rubber band design under the bust wide bottom, partition design before and after the shoulder, built-thin glass design on the chest, three times partition and hollowing outdesign on the back. This thesis testifid the artificial network model and zoning design theory of sport brassieres by dressing experiments.There was a shortage about the experimental ones that thermal-wet comfort under bust was discontented,but the pressure value and the thermal-wet comfort of other parts were superior to similar sport brassieres in the market.