摘要
为了确保基于蒙特卡罗的接地网支路可测性分析结果的正确率并提高分析速度,提出一种改进方法。在对一批随机生成的各条支路电阻取值样本分别进行仿真故障诊断的基础上,采用最小生成树算法对明晰与不确定支路进行自动判别。探讨了补充明确顶点、迭代初值调整、支路二次分类、相近样本调整和理想样本自动生成等改进措施。以一个60条支路的接地网为例,详细说明了所建议方法并分析了各项措施的必要性和可行性,结果表明:所提出的各项改进措施都是可行的和有效的,并且能够避免人为选取阈值对支路可测性分析带来的影响、有效地提高判别正确率和减少判别处理时间。
To guarantee the validity of the Monte-Carlo based testability evaluation of grounding grids corrosion diagnosis, a refined approach including several improvements is presented. Based on the simulated corrosion diagnosis results of a group of randomly formed branch resistances, the clear-branches and uncertain-branches are identified according to a minimum spanning tree based algorithm. The improvement method of auxiliary clear vertex and unknown vertex adding, iteration initial values adjustment, second classification of branches, similar sample adjustment and the ideal sample automatic generation are described in detail. A grounding grid with sixty branches is used as an example. The analysis results show that above improvement approaches are necessary and feasible. The influence of threshold setting by experience on the validity of testability evaluation is avoided. Besides, the efficiency of testability evaluation process is remarkably increased.
出处
《电测与仪表》
北大核心
2009年第12期9-13,24,共6页
Electrical Measurement & Instrumentation
基金
新世纪优秀人才支持计划资助项目(NCET050875)
关键词
接地网
故障诊断
蒙特卡罗模拟
可测性
最小生成树算法
grounding grids, corrosion diagnosis, Monte-Carlo approach, testability, minimum spanning tree based algorithm