基于多元线性回归的变电设备状态评价权重修正模型

Weight Correction Model of Condition Assessment for Substation Equipment Based on Multiple Linear Regression Method

为解决多数电网公司所采用的基于导则的打分制变电设备状态评估方法中缺少针对权重修正的闭环反馈,致使评估权重较为固定的问题,文中引入变电设备状态评价权重修正模型,基于历史评估数据与经与状态评级事后有效性评价得到的修正结果,通过多元线性回归模型,给出回归后的理论修正权重,并可根据该结果对状态量权重进行修正。通过理论及应用实例验证表明:文中所提出的通过结合客观的历史检修、运维、评估数据与事后有效性验证的权重修正模型,是一种较为有效的权重修正方法,可以为更准确地为评估变电设备的状态情况提供参考。

Condition assessment methods of substation equipment based on scoring guidelines, which are adopted by most power grid companies, lacks of feedback loop for weight correction. As a result, the weight of assessment is relatively fixed, and could not form a scientific and rational correction scheme. In this paper, the authors introduce a condition assessment weight correction model for substation equipment based on the multiple linear regression model. By testing the theory and application examples, it shows that the weight correction model which combines objective historical overhaul, maintenance, data of condition assessment and assessment validation is a more effective weight correction method, and it can provide reference for more accurate condition assessment of substation equipment.