一种分布式PLC边缘计算系统及数据通信方法
A Distributed PLC Edge Computing System and Data Communication Method
邢汉 1熊光亚 1刘林兴 1霍瑞1
作者信息
- 1. 国网电力科学研究院/南瑞集团公司,江苏省南京市 210003
- 折叠
摘要
本文介绍了一种分布式PLC边缘计算系统及其数据通信与处理方法:包括将系统中PLC运行相关的所有计算任务按照PLC主体功能的类型,划分为各个相对独立的边缘计算任务,由分布在不同PLC中的边缘计算模件承担;每种类型的边缘计算任务均包含实时数据获取、数据预处理、特征值计算、告警判别与输出几项基本任务;CPU模件和边缘计算模件采用独立的高速双向接口进行交互,边缘计算模件对从CPU模件接收的实时数据进行预处理后计算出特征值,并比较原始数据和特征值之间的差异程度,判断是否输出告警信息;各个边缘计算模件之间还包括联动计算告警功能.将数据处理和分析的过程放在离用户和设备更近的位置,减少了数据传输的延迟时间,从而提高了响应速度;将一部分计算任务从数据中心转移到边缘计算装置上处理,可以减少与数据中心之间的数据传输量,从而减轻网络负载;执行一些本地的决策和控制任务,不必完全依赖于数据中心,可以实现更快速地故障处理和更精确的设备控制;边缘功能可以在无网络连接的情况下继续运行和处理数据,具备更好的离线支持能力;实现了整个系统数据处理分析任务的去中心化,显著降低了中心单一设备或软件崩溃对系统运行造成的风险.
Abstract
This paper introduces a distributed PLC edge computing system and its data communication and processing methods:including dividing all computing tasks related to PLC operation in the system into relatively independent edge computing tasks according to the type of PLC main function,which are undertaken by edge computing modules distributed in different PLCs;Each type of edge computing task includes several basic tasks:real-time data acquisition,data pre-processing,eigenvalue calculation,alarm discrimination and output;The CPU module and edge computing module use independent high-speed two-way interface for interaction.The edge computing module preprocesses the real-time data received from the CPU module to calculate the eigenvalue,compares the difference between the original data and the eigenvalue,and judges whether to output alarm information;Each edge computing module also includes the linkage calculation alarm function.Placing the process of data processing and analysis closer to users and devices reduces the delay time of data transmission,thereby improving response speed;Transferring some computing tasks from the data center to edge computing devices for processing can reduce the amount of data transmission with the data center,thus reducing the network load;Performing some local decision-making and control tasks without completely relying on the data center can achieve faster fault handling and more precise equipment control;The edge function can continue to run and process data without network connection,with better offline support capabilities;Realized the decentralization of data processing and analysis tasks for the entire system,significantly reducing the risk of system operation caused by single device or software crashes in the center.
关键词
边缘计算/特征值/联动/去中心化Key words
edge computing/characteristic value/linkage/out of center引用本文复制引用
基金项目
国家电网有限公司总部科技项目(5700-202140381A-0-0-00)
出版年
2024
NETL
NSTL
万方数据