一种基于关联规则的变压器关键参量提取方法

Research on the Extraction Method of Key Parameters of Transformer Based on Association Rules

从众多的指标中筛选关键参量可提高变压器状态识别的准确性和效率。针对变压器的状态识别,提出一种基于关联规则的变压器状态关键参量提取方法。首先在分析变压器状态指标参量与故障类别关联关系的基础上,构建较为全面的变压器基础参量体系,然后运用关联规则将变压器故障类型与指标参量之间的关系进行量化,通过计算得到各指标参量的支持度与置信度,以满足两者最小阈值为条件提取出变压器状态关键参量,从而构建变压器状态关键参量体系。最后以某变电站故障统计数据为例进行验证,为变压器监测数据分析提供一种新的思路。

The selection of key parameters among many factors can help improve the accuracy and efficiency of transformer state evaluations. According to the transformer status evaluation, this paper proposed a method for extracting key parameters of transformer based on association rules. Firstly, the basic parameter system of transformers was established based on analyzing the association between the parameters of the transformer status and the fault symptoms. Then, association rules were adopted to quantify the relationship between the fault symptoms and the parameters of transformer status. By computation, support degree and confidence degree of various parameter indexes were obtained to extract key parameters of the transformer status under minimum threshold.Then, the system of key parameters of the transformer status was established. Finally, an example based on statistical components＇ failure data of the substation was verified, which offered a new method for processing data gained from monitoring transformers.