节点文献

基于AI芯片的电力边缘智能终端:结构框架及其应用场景

Power Edge Intelligent Terminal Based on AI Chip: Structural Framework and Application Scenarios

  • 推荐 CAJ下载
  • PDF下载
  • 不支持迅雷等下载工具,请取消加速工具后下载。

【作者】 周胤宇王波朱丹蕾马富齐罗鹏王红霞

【Author】 ZHOU Yinyu;WANG Bo;ZHU Danlei;MA Fuqi;LUO Peng;WANG Hongxia;School of Electrical Engineering and Automation, Wuhan University;Central Southern China Electric Power Design Institute Co., Ltd., China Power Engineering Consulting Group;

【通讯作者】 王波;

【机构】 武汉大学电气与自动化学院中国电力工程顾问集团中南电力设计院有限公司

【摘要】 随着电力物联网发展及能源数字化转型,电力边缘智能终端的适用场景不断拓展。AI芯片由于其本身智能分析及并行计算能力,成为电力边缘终端的主流核心器件之一,如何通过系统性的分析,根据应用场景选择合适的AI芯片成为目前亟需解决的问题。文章首先对AI芯片的技术架构和优缺点进行对比分析,然后根据电力系统的智能感知要求提出电力边缘智能终端的结构框架,并详细分析边缘智能终端构建的关键技术,最后结合电力典型应用场景提出了国产化可替代的AI芯片定制化选型方案。文章所提AI芯片化的电力边缘智能终端可以为电力边缘终端的设计、研发及应用提供指导,并具有普适性的理论和实用价值。

【Abstract】 With the development of the power Internet of things and the transformation of energy digitization,the application scenario of the power edge smart terminal is constantly expanding.AI chip has become one of the mainstream core devices of power edge terminal due to its intelligent analysis and parallel computing capability.However,how to select appropriate AI chip according to the application scenario through systematic analysis has become an urgent problem to be solved.This paper first analyzes the AI chip technology architectures and their advantages and disadvantages,the structural framework of the edge smart terminal is proposed,and the key technologies for the construction of the edge smart terminal are analyzed in detail.Finally,the customized selection scheme of the domestic alternative AI chip is proposed based on the typical power application scenarios.The AI chip power edge smart terminal proposed in this paper can effectively provide guidance for the design,development and application of power edge terminals,and has universal theoretical and practical value.

  • 【文献出处】 电力信息与通信技术 ,Electric Power Information and Communication Technology , 编辑部邮箱 ,2021年09期
  • 【分类号】TM73
  • 【被引频次】3
  • 【下载频次】542
节点文献中: 

本文链接的文献网络图示:

本文的引文网络