...
首页> 外文期刊>The Journal of the Textile Institute. 1, Fibre Science and Textile Technology >A Study on the Capability of a Neural Network Ranking Fibre Parameters Having an Influence on Yarn Properties
【24h】

A Study on the Capability of a Neural Network Ranking Fibre Parameters Having an Influence on Yarn Properties

机译:影响纤维性能的神经网络对纤维参数排名的能力研究

获取原文
获取原文并翻译 | 示例
           

摘要

Though artificial neural networks have been used in a wide range of textile-related problems, the potential of the networks for discerning the relative importance of the input parameters has never been tested. Nor does there exist any reliable technique for this purpose. In the work reported in the present paper, neural networks were first created for predicting ring-spun-cotton-yarn properties from a set of fibre properties measured on high-volume instruments (HVI). The relative importance of the input parameters was evaluated by using two different methods, namely, 'skeletonozation', a method reported in the literature for pruning neural networks, and a newly developed method based on first-order sensitivity analysis. The sensitivity analysis was found to work significantly better.
机译:尽管人工神经网络已被广泛用于与纺织品有关的问题,但从未检验过网络识别输入参数的相对重要性的潜力。也没有为此目的提供任何可靠的技术。在本文报道的工作中,首先创建了神经网络,用于从在大容量仪器(HVI)上测量的一组纤维特性预测环锭纺棉纱的特性。输入参数的相对重要性通过使用两种不同的方法进行评估,即“骨架氧化”,文献中报道的用于修剪神经网络的一种方法以及一种基于一阶敏感性分析的新开发方法。发现灵敏度分析的效果明显更好。

著录项

相似文献

  • 外文文献
  • 中文文献
  • 专利
获取原文

客服邮箱:kefu@zhangqiaokeyan.com

京公网安备:11010802029741号 ICP备案号:京ICP备15016152号-6 六维联合信息科技 (北京) 有限公司©版权所有
  • 客服微信

  • 服务号