高压电网配电设备故障智能诊断系统的设计  

Partition Mining Technology in Distribution System Fault Diagnosis of Application

作　　者：冯明卿[1] 冯玉东[1] 马斌[2] 

机构地区：[1]郑州电力高等专科学校,郑州450004 [2]华北水利水电学院,郑州450011

出　　处：《科技通报》2013年第2期148-150,共3页Bulletin of Science and Technology

基　　金：2011年河南省教育厅科研项目(2011C510007)

摘　　要：针对在实际的高压电网配电系统中,高压环境较为危险,故障信息特征极难获得,信息不完整,导致无法有效及时地寻找故障原因的问题。本文将智能无限传感器网络的思想引入到配电系统的故障诊断过程中,将保护、继路器的状态作为系统进行故障检测的条件属性,故障发生的器件或者区域作为决策结果属性。系统的软件设计中,利用相关性反馈与交互式的挖掘算法结合,对供电系统中的各种故障信息进行决策的推理与诊断。实验结果表明,该故障诊断系统在一定的高压电网配电规模和保护动作不完整的情形下诊断的准确率达到了77%。In the actual high voltage grid for distribution systems,high-pressure environment is more dangerous,the failure information features extremely difficult to obtain,information is not complete and lead to not effective in the search for the cause of the problem.The smart infinite sensor network into the idea of distribution system fault diagnosis of process.Will protect the state of method,as system fault detection condition attribute,the device failure or area as the decision results attributes.In the design of the system software,using correlation with interactive feedback algorithm for mining combination of the power supply system,in various fault information to make decisions and diagnosis reasoning.The experimental results show that the fault diagnosis system in a certain high voltage grid power distribution scale and protection movement incomplete in diagnosis accuracy under 77%.

关 键 词：配电系统 故障诊断 关联规则 

分 类 号：TP137[自动化与计算机技术—控制理论与控制工程]