基于参数辨识的变压器绕组变形在线监测方法

Method of On-line Monitoring of Transformer Winding Deformation Based on Parameter Identification

变压器绕组变形导致的内部匝间故障已经上升至事故率的首位,对电网安全性危害极大。为此,提出一种变压器绕组变形在线监测工程方法。利用以傅里叶为理论基础的动态向量模型,在保持高精度的条件下简化变压器电磁暂态模型,并讨论不同运行状态下模型参数辨识解的个数。当变压器空载合闸时辨识出绕组电阻参数,在变压器稳态运行时利用已辨识出的电阻参数修改参数辨识模型,并辨识漏感参数。利用绕组变形与漏电感参数之间的关系,实现对变压器绕组变形在监测。建立了变压器绕组变形的仿真模型,使用有限元方法计算变压器绕组变形的短路电感值,验证了变压器绕组在线监测方法的有效性。

Internal turn-to-turn faults caused by transformer winding deformation have become the most common accident, which does serious damage to power grid security. An effective engineering method of on-line monitoring of transformer winding deformation is proposed in this paper. The dynamic vector model based on Fourier theory was applied to simplify the electromagnetic transient model of transformers while maintaining its high accuracy, and the number of identification parameters of model under different operational states was discussed. The winding resistance parameters were identified when the transformer closed to the unloaded transmission line, and were applied to modify the identification model, and then leakage inductance parameters were identified. Transformer monitored online due to the deformation and leakage winding deformation can be relationship between winding inductance parameters. The simulation model of transformer winding deformation was constructed and the short-circuit inductances during transformer winding deformation were calculated by finite element method. The effectivity of the method for transformer winding on-line monitoring was validated.