基于深度学习的调度自动化报文合规性研究与实现  

Research and Implementation of Dispatching Automation Message Compliance Based on Deep Learning

作　　者：朱文[1] 方文崇[1] 李金[1] 谢型浪 谢虎 ZHU Wen;FANG Wen-chong;LI Jin;XIE Xing-lang;XIE Hu(China Southern Power Grid Co.,Guangzhou 510770 China;Digital Grid Research Institute,China Southern Power Grid.,Guangzhou 510770 China)

机构地区：[1]中国南方电网有限责任公司,广东广州510770 [2]南方电网数字电网研究院有限公司,广东广州510770

出　　处：《自动化技术与应用》2021年第2期156-159,共4页Techniques of Automation and Applications

摘　　要：本文基于大数据及机器学习,通过分析调度自动化系统特定规约的下行控制报文,找出自动控制及人工控制的报文合规性,形成合规专家库;对海量的典型调度自动化系统主站和厂站间的交互控制报文进行关联性识别,基于大数据及深度机器学习,分析101、104等调度自动化系统规约报文在时间、控制频率、电网状态关联性、控制范围等方面的规律,形成调度自动化系统控制功能的报文合规专家库,并通过模拟试验并且同时对CIM模型、操作票进行解释,对异常报文进行告警。Based on big data and machine learning,this paper analyzes the downlink control message of the specific specification of the dispatching automation system,finds out the message compliance of automatic control and manual control,and forms the compliance expert database;identifies the association of the interaction control message between the main station and the plant station of the massive typical dispatching automation system,and analyzes the 101,104 and other adjustments based on big data and deep machine learning.The rules of the protocol message in time,control frequency,power grid state relevance and control range of the automation system are formed to form the message compliance expert database of the control function of the dispatching automation system,and the abnormal message is alarmed through simulation test and interpretation of CIM model and operation ticket at the same time.

关 键 词：调度自动化 大数据 机器学习 控制 CIM模型 

分 类 号：TP311.13[自动化与计算机技术—计算机软件与理论]