人工智能处理在自动化变电站保护系统中的实时故障诊断与评估  

Real-time fault diagnosis and evaluation of artificial intelligence processing in automated substation protection system

作　　者：李扬笛 林爽 郑州 黄建业 钱健 杨彦 LI Yangdi;LIN Shuang;ZHENG Zhou;HUANG Jianye;QIAN Jian;YANG Yan(State Grid Fujian Electric Power Co.,Ltd.,Electric Power Science Research Institute,Fuzhou 350007,China)

机构地区：[1]国网福建省电力有限公司电力科学研究院,福州350007

出　　处：《自动化与仪器仪表》2024年第9期201-205,210,共6页Automation & Instrumentation

基　　金：国家电网项目,福建电科院数字中心2023年基于人工智能平台的AI资源和运营价值评估服务(172354618)。

摘　　要：由于自动化变电站保护系统故障诊断效率低、精度差的问题频频发生。为此,提出基于人工智能的变电站设备故障综合诊断与评估方法。首先,通过主分量分析技术对电力信号降维,结合支持向量机方法分类电力保护设备数据组。利用可达矩阵分析信息在源宿之间的可达性,确定自动化变电站保护系统失效树,定性分析故障引起的不可达信息,将检测得到的差异化数据特征数集进行归一处理,判断设备是否处于故障状态并且评估风险。实验结果表明:所提检测方法能够精准诊断保护设备的电压互感器、电流互感器、断路器、隔离开关所有处于故障状态的设备位置。自动化变电站保护设备故障风险的准确率约为88.94%。评估时间仅在0.194 s~0.237 s之间,所提方法能确保整个变电站电力系统的安全运行,在电力系统中具有较高的应用价值。Due to the frequent occurrence of low efficiency and poor accuracy in fault diagnosis of automatic substation protection systems,a comprehensive diagnosis and evaluation method for substation equipment faults based on artificial intelligence is proposed.Firstly,the principal component analysis technique is used to reduce the dimensionality of the power signal,and the support vector machine method is used to classify the data group of power protection equipment.Using the reachability matrix to analyze the accessibility of information between source and destination,determine the fault tree of the automation substation protection system,qualitatively analyze the unreachable information caused by faults,normalize the differential data feature number set detected,determine whether the equipment is in a fault state,and evaluate the risk.The experimental results show that the proposed detection method can accurately diagnose the position of all equipment in fault state,such as voltage transformers,current transformers,circuit breakers,and isolation switches,in the protection equipment.The accuracy rate of fault risk for protection equipment in automated substations is approximately 88.94%.The evaluation time is only between 0.194 seconds and 0.237 seconds,and the proposed method can ensure the safe operation of the entire substation power system,which has high application value in the power system.

关 键 词：自动化变电站 信号降维 支持向量机 变电站保护设备 故障诊断 

分 类 号：TP274[自动化与计算机技术—检测技术与自动化装置]