灰色定权聚类和变权模式在变压器状态评估中的应用研究

针对变压器状态评估因素具有模糊性与灰色性的特点,定权重不能准确评估变压器状态等问题,提出一种基于灰色定权聚类和变权模式的状态评估方法。应用该方法逐层评估变压器状态。首先,引入关联规则中的置信度确定指标定权重矩阵,并对指标进行统一化处理;根据变压器状态等级划分,建立三角白化权函数,利用灰色聚类得到聚类系数矩阵和故障层的状态评估结果;然后,利用变权综合模式获得变权重系数,进而采用模糊评判法评估变压器综合状态。以变压器试验报告数据为依据进行实例分析,结果表明,所提出的变压器状态综合评估方法能够客观、准确地评估变压器状态。

Application of grey fixed weight clustering and variable weight model in transformer condition evaluation

Aiming at the problem that the transformer condition evaluation factors have fuzzy and grey characteristics and fixed weight (FW) can not accurately evaluate the condition,we proposed a condition evaluation method for transformer based on grey fixed weight clustering and variable weight model.The method is applied to evaluate transformer condition layer by layer.Firstly,the confidence degree of the association rules is introduced to determine the FW of key indicators.According to the classification of transformer state grade,the triangle whiten function (TWF) is established.The grey clustering coefficient (GCC) matrix and fault layer evaluation results are obtained through grey clustering (GC) method.Then the variable weight (VW) is obtained by combining variable weight synthesis mode,and then the transformer condition is evaluated by fuzzy evaluation method.The transformer test report data is carried out as the case analysis and results show that the method we proposed can assess the condition of the transformer objectively and accurately.