基于GC-FD分析及GM(1,N)模型预测的变压器故障监测与保护  

Monitoring and Protection against Transformer Failures Based on GC-FD Analysis and GM(1,N) Model Forecast

作　　者：陈冬霞[1] 石辉[2] 

机构地区：[1]广东电网公司东莞供电局,广东东莞523008 [2]湖南省电力公司调度通信局,湖南长沙410007

出　　处：《广东电力》2011年第9期36-40,共5页Guangdong Electric Power

摘　　要：提出基于灰关联(grey connection,GC)-模糊距离(fuzzy distance,FD)综合分析以及GM(1,N)模型的色谱气体浓度预测的变压器故障监测与保护体系的实现方法。根据变压器在线监测的现状和实时保护的需要,提出变压器故障前主动保护的思路。在GM(1,N)建模获取色谱气体近期预测数据的基础上,综合静态灰关联分析、动态灰关联分析和模糊距离分析,提出把综合关联系数用于变压器故障准确预报,并将试调整数据回代GM(1,N)模型进行故障再预测,确定应对故障的调整措施和数据。该体系能准确可靠地预报并帮助解除变压器故障,适于软件实现并配合变压器主保护,算例分析表明了体系的实用性和有效性。The paper brings forward implementation approach for monitoring and protection system against transformer failures based on grey connection （GC）-fuzzy distance （FD） integrated analysis and chromatographic gas concentration forecast of GM （i, N） model. In accordance with current status of on-line monitoring of transformers and the demand of real-time protection, the paper suggests to protect transformer before occurrence of failures. On the basis of modeling of GM （1, N） to obtain recent chromatographic gas forecast date of gas and in combination with static grey connection analysis, dynamic grey connection analysis and fuzzy distance analysis, the paper proposes that comprehensive association coefficient be used for correct forecast on transformer failures and data of beta testing be substituted back to GM （1, N） model for failure reforecast so as to determine adjusting measures and data for failure handling. The system can accurately and reliably forecast and help eliminate transformer failures and it is applicable to software implementation cooperating with main protection of transformers. The analysis of calculation example shows practicability and effectiveness of this system.

关 键 词：灰关联-模糊距离综合分析 色谱气体浓度预测 GM(1 N)模型 主动保护 

分 类 号：TM762[电气工程—电力系统及自动化] TM407