摘要
针对当前变压器数量庞大、状态量数据众多,无法准确掌握变压器运行状态等管理现状,设计开发了一种基于贝叶斯网络的群组变压器健康状况分析系统。该系统基于虚拟机技术,整合所有可获取数据信息,按照影响的部件或运行模式建立贝叶斯条件概率拓扑网络,通过信息分类开展故障影响因素诊断分析,结合不同状态信息的权重因子,得到变压器评估单元故障风险的分值,合成设备故障风险概率分值。并引入变压器重要度变量绘制群组变压器运行状态风险等级图,根据变压器落入的颜色区域实现群组设备的分级管理。系统在变压器故障风险准确预警和诊断效率方面都有显著优势,可帮助电力企业建立明晰、高效的群组变压器检修计划,提高设备利用率,实现设备效益最大化的管理目标。
In view of the large number of transformers and the large amount of state data,it is impossible to accurately grasp the current management status of transformer operation,etc.A group transformers health analysis system based on Bayesian network is designed and developed.Based on virtual machine technology,the system integrats all available data and information,establishes Bayesian conditional probability topology network according to the influenced component or running modes,carries out the fault influencing factor diagnosis and analysis through the information classification,obtains the score of transformer evaluation unit failure risk combined with the weight factors of different state information,and synthesizes the probability score of equipment failure risk.In addition,the transformer importance variable is introduced to draw the group transformers operation risk level diagram,and the hierarchical management of the group equipment is realized according to the color area that the transformer falls into.The system has significant advantages in accurate early warning of transformer failure risks and diagnostic efficiency,which can help power companies establish a clear and efficient group transformers maintenance plan,improve equipment utilization,and achieve the management goal of maximizing equipment benefits.
作者
杨玥
康琪
杨军
姜涛
YANG Yue;KANG Qi;YANG Jun;JIANG Tao(Inner Mongolia Electric Power Science&Research Institute Branch,Inner Mongolia Power(Group)Co.,Ltd.,Hohhot 010020,China)
出处
《测控技术》
2021年第9期38-47,共10页
Measurement & Control Technology
基金
2020年内蒙古电力(集团)有限责任公司重点科技项目(2020-08)。
关键词
群组变压器
贝叶斯网络
交叉熵
缺失数据处理
分级诊断
group transformers
Bayesian network
the cross entropy
missing data processing
hierarchical diagnosis
