数字孪生模型下油浸式变压器内部故障检测方法

Internal Fault Detection Method of Oil-immersed Transformer Based on Digital Twin Model

针对因油浸式变压器内部动态特征难以提取,并因噪声干扰,导致故障检测误差较大问题,提出一种基于数字孪生模型的故障检测方法来实现有效解决。构建油浸式变压器动力学模型,分析故障主要成因,将内部绝缘体的失效时间当作线性指标,采用威布尔分布函数求解故障发生概率与失效时间的线性关联。建立变压器故障停留时间、运行天数以及故障概率方差间拟合函数,得到的拟合值作为故障检测模型的初始数据参照,利用数字孪生模型,根据变压器数据集给出几何、行为、物理以及规则的检测尺度,不同尺度代表变压器的不同信息,建立动态故障概率计算函数,将现场数据代入计算得到故障检测阈值,再对比数据完成检测。实验数据证明,所提方法检测故障精准度高,在多种环境下均能保证检测结果,并具有较强的鲁棒性。

Aiming at the problem that the internal dynamic characteristics of oil-immersed transformer are difficult to extract,and the noise interference leads to large fault detection error,a fault detection method based on digital twin model is proposed to effectively solve the problem.Build the dynamic model of oil-immersed transformer,analyze the main causes of faults,take the failure time of internal insulator as a linear index,and use Weibull distribution function to solve the linear relationship between the failure probability and failure time.Establish the fitting function between transformer fault dwell time,operation days and fault probability variance.The fitting value obtained is used as the initial data reference of the fault detection model.The digital twin model is used to give the geometric,behavioral,physical and rule detection scales according to the transformer data set.Different scales represent different information of the transformer,and establish the dynamic fault probability calculation function,substitute the field data into the calculated fault detection threshold,and then compare the data to complete the detection.The experimental data shows that the proposed method has high accuracy in fault detection,can guarantee the detection results in a variety of environments,and has strong robustness.