经编提花鞋材的自动排版及CUDA的并行实现  

Automatic layout and CUDA parallel implementation of warp knitted jacquard shoe upper

作　　者：吴妙缘 张爱军 张燕婷 WU Miaoyuan;ZHANG Aijun;ZHANG Yanting(Engineering Research Center for Knitting Technology,Ministry of Education,Jiangnan University,Wuxi 214122,China)

机构地区：[1]江南大学针织技术教育部工程研究中心,江苏无锡214122

出　　处：《丝绸》2021年第12期126-132,共7页Journal of Silk

基　　金：中央高校基本科研业务费专项资金重点项目(JUSRP51727A);泰山产业领军人才项目(tscy20180224)。

摘　　要：经编提花鞋材具有良好的性能,被广泛应用于运动鞋的鞋面。鞋材产品开发时为提高面料利用率工艺人员需对鞋材进行排版,但手工排版效率低下。为了实现经编提花鞋材的快速精准排版和提高面料的利用率,文章介绍了3种常见的鞋材排版方式,给出了基于边界碰撞检测的排版算法。根据排版算法可并行的特征,基于统一计算设备构架(CUDA)实现了算法。实验结果表明,CUDA并行处理的执行速度明显快于CPU串行处理,这一算法实现了鞋材排版的快速化与自动化,同时鞋材位置的精准放置也使织物利用率得到了提升。The warp-knitted jacquard shoe upper has been widely used in the upper of sports shoes due to its excellent performance.In order to improve fabric utilization during shoe product development,shoe upper layout needs to be conducted by technology personnel,but manual layout is inefficient.In order to achieve fast and accurate layout of warp-knitted jacquard shoe uppers and improve fabric utilization,three common shoe layout methods were introduced in this paper,and a layout algorithm based on boundary collision detection was provided.According to the parallel feature of the layout algorithm,the algorithm was implemented based on CUDA.The experiment results show that the speed of parallel execution of CUDA is significantly faster than that of CPU serial execution.The algorithm has realized the high-speed and intelligent layout of shoe uppers.At the same time,the precise placement of shoe upper has also improved fabric utilization.

关 键 词：经编鞋材 贾卡提花 排版 面料利用率 并行算法 CUDA 

分 类 号：TS943.68[轻工技术与工程—服装设计与工程]