...
首页> 外文期刊>ThyssenKrupp Magazine >Top precision through artificial intelligence
【24h】

Top precision through artificial intelligence

机译:人工智能带来的最高精确度

获取原文
获取原文并翻译 | 示例
           

摘要

TAKO - an intriguing word. But what does this cryptic term used at ThyssenKrupp Steel in Duisburg really mean? An attempt to unveil the secret takes us into a scientific world that was developed and implemented by U.S. researchers (such as John von Neumann and later Donald 0. Hebb and Marvin Minsky) exactly 61 years ago: Artificial neural networks. TAKO stands for "a combination of pickle and tandem mill." A rather dull phrase compared to the way in which neural networks increasingly mirror the function of human nerve cells. Neural networks are state-of-the-art computer programs. For the moment, however, software experts estimate that their actual performance still amounts to no more than that of a fly's brain. Still, "an analogy exists between human and technical comprehension in our TAKO processes in the context of the neural system," says Bert Freygang, engineer and head of operations for technology at ThyssenKrupp Steel's Duisburg plant in the automotive division. "Within the neural network, I consider the hard figures and the data that flow into the computer systems via the communication levels particularly relevant. The neural network has to process these signals and respond with specific signals." Dr. Karl-Ulrich Koehler, the chairman of the management board of ThyssenKrupp Stahl AG, summed up the key benefits at the inauguration of TAKO three years ago: "In process automation, artificial neural networks are used to complement mathe- matical models. This makes it possible to balance out influencing variables that are otherwise difficult to grasp, such as the current status of the mill. Before the start of operations, the neural networks were trained in the laboratory."
机译:TAKO-一个有趣的词。但是在杜伊斯堡的蒂森克虏伯钢铁公司使用的这个神秘术语到底意味着什么?揭开这个秘密的企图将我们带入了一个完全由61岁的美国研究人员(例如John von Neumann和后来的Donald 0. Hebb和Marvin Minsky)开发和实施的科学世界:人工神经网络。 TAKO代表“酸洗和串磨的组合”。与神经网络越来越多地反映人类神经细胞功能的方式相比,这是一个相当呆板的短语。神经网络是最先进的计算机程序。然而,目前,软件专家估计他们的实际性能仍然不超过苍蝇的大脑。蒂森克虏伯钢铁公司汽车部门杜伊斯堡工厂的工程师兼技术主管Bert Freygang表示:“在我们的TAKO过程中,在神经系统的背景下,人类和技术理解之间仍然存在类比。” “在神经网络中,我认为通过通信级别流入计算机系统的硬数字和数据特别相关。神经网络必须处理这些信号并以特定信号做出响应。”蒂森克虏伯·斯塔尔股份公司(ThyssenKrupp Stahl AG)管理委员会主席Karl-Ulrich Koehler博士总结了三年前TAKO就职典礼的主要好处:“在过程自动化中,人工神经网络用于补充数学模型。这样就可以平衡原本很难掌握的影响变量,例如工厂的当前状态。在开始运行之前,神经网络已经在实验室进行了训练。”

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