数字孪生变电站框架设计与关键技术研究  

Framework Design and Key Technology Research of Digital Twin Substation

作　　者：张冀[1,2,3] 马也 张荣华 朵春红[1,2] ZHANG Ji;MA Ye;ZHANG Ronghua;DUO Chunhong(School of Control and Computer Eng.,North China Electric Power Univ.,Baoding 071003,China;Eng.Research Center of Intelligent Computing for Complex Energy Systems,Ministry of Education,Baoding 071003,China;Hebei Key Lab.of Knowledge Computing for Energy&Power,Baoding 071003,China)

机构地区：[1]华北电力大学计算机系,河北保定071003 [2]复杂能源系统智能计算教育部工程研究中心,河北保定071003 [3]河北省能源电力知识计算重点实验室,河北保定071003

出　　处：《工程科学与技术》2023年第6期15-30,共16页Advanced Engineering Sciences

基　　金：河北省省级科技计划资助(22310302D)。

摘　　要：数字孪生(DT)技术的发展给电力系统的智能化管理带来诸多便利,然而作为推动能源电力行业数字化、智能化的关键技术,相关研究与应用还处于初期起步阶段,亟需开展系统性研究,以突破适应能源电力行业特殊性的数字孪生关键技术。为此,本文以变电站作为应用场景,首先对变电站运维管理现状进行了分析,指出现阶段存在变电站巡检模式缺陷、模型精度不够、感知参量单一、缺乏对数据的深度挖掘等问题。以此设计了一套涵盖空间信息与设备建模、变电主设备机理建模、智能反馈控制、设备感知网络、基于数据驱动的数字孪生体模型的仿真,以及3维可视化渲染与应用的数字孪生变电站模型框架。然后讨论了变电站数字孪生技术面临的问题与挑战:专业智能研制方面,变电站数字孪生系统对传感器的感知精度、智慧功能集成度、抗干扰能力、功耗以及供能方式等提出更高要求;海量数据存储与计算方面,海量多源异构感知数据如何高效存储与利用,以及如何优化算力资源分配,满足系统实时计算需求;模型研究方面,现有数据驱动模型精度不足,模型不可解释且缺乏论证,影响整个系统的可靠运行;数据方面,如何解决“数据安全”与“数据孤岛”问题;3维虚拟实体构建与动态更新方面,如何低成本快速构建可用的高精度3维变电站模型,以及研究设计3维模型的动态更新算法,以保证孪生系统实时空一致性。进而探讨了解决问题的关键技术:通过感知理论研究与提升制作工艺、结合分布式云存储与云边协同计算、数据驱动与机理知识融合建模、可信联邦学习等先进技术,以及实景3维重建与点云动态可视化等相关技术,在满足隐私保护和数据安全前提下,建立“形”与“态”相融合的变电站数字孪生系统,实现变电站运行状态全感知、全生命周期数据智能管理与The development of digital twin(DT)technology brings many conveniences to the intelligent management of power systems.However,as a key technology to promote the digitalization and intelligence of the energy and power industry,related research and applications are still in the initial stage.Systematic research is urgently needed to break through the adaptation key technologies of digital twin engineering application for the particularity of the energy and power industry.For this reason,the substation was took as an application scenario in this paper,the current situation of the operation and maintenance management of the substation were firstly analyzed,and the defects of the substation inspection mode,the insufficient model accuracy,the single perception parameter as well as the lack of in-depth data mining were pointed out.Then,a digital twin substation model framework was designed,which includes spatial information and equipment modeling,main equipment mechanism modeling,intelligent feedback control,equipment sensing network,model simulation and 3D visualization rendering and applications.The problems and challenges of substation digital twin technology are discussed,including professional intelligent development,mass data storage computing resource allocation,model research,data security,3D virtual entity construction and dynamic update.In view of the problems and challenges raised,the research on key technologies to solve the problems was discussed,including the development of new sensors,distributed storage,cloud-edge collaborative computing,fusion modeling,trusted federation learning and other advanced technologies,as well as the 3D reconstruction of real scene and the dynamic visualization of point cloud.Combined with the design of digital twin model and the discussion of key technical issues,a set of application-oriented digital twin substation system design schemes were given,focusing on the real-time monitoring of equipment in the station,equipment fault diagnosis and fault prediction,operation and main

关 键 词：变电站 数字孪生 框架设计 智能巡检 压缩感知 融合建模 可信联邦学习 3维可视化 

分 类 号：TM715[电气工程—电力系统及自动化]