电工技术学报2025，Vol.40Issue(5)：1639-1651,13.DOI:10.19595/j.cnki.1000-6753.tces.240264
基于简支绕组横梁算法的变压器绕组形变矢量化监测模型
Vectorization Monitoring Model of Transformer Winding Deformation Based on Simply Supported Winding Beam Algorithm
摘要
Abstract
Winding deformation is one of the important causes of transformer failure and is the focus of transformer condition monitoring.To achieve accurate monitoring of the direction and degree of transformer winding deformation,this paper combines the principle of a simple beam with the Brillouin optical time-domain analysis technique and proposes a vectorized monitoring algorithm for transformer winding deformation. Firstly,based on the theoretical derivation of material mechanics,the Brillouin frequency shift of axial deformation is proportional to the distance from the optical fiber to the neutral axis,and the Brillouin frequency shift of radial deformation is equal to the law,and the relationship between the Brillouin frequency shift and its axial and radial components is established,and the Brillouin optical time-domain analysis is combined to establish a vectorized monitoring model of the deformation of transformer windings that can be used in actual operation. The vectorized fiber optic sensors are designed,and the fiber optic composite conductor model is built on a single conductor.5 sets of axial and radial deformations of different degrees are applied by a high-precision tensile machine,and the deformation monitoring is completed with the help of the Brillouin optical time-domain reflection technology,and the result verifies that the conductor is subjected to the bulging and upward deviation as the positive strain,and the depression and downward deviation as the negative strain,and the coordinate system of the direction of deformation is constructed,and the deformation direction determination can be achieved by the demodulation result.The deformation direction can be determined by the demodulation results,calculating the experimental error of the conductor axial deformation monitoring error average value of 6.60%,radial deformation monitoring error average value of 5.88%,the analysis of the error caused by the high-precision press slight jitter and uneven application of the epoxy resin adhesive,but demodulation effect is generally good,to achieve the deformation of the fiber optic composite conductor direction and degree of monitoring. To effectively prevent the fiber optic offset caused by vibration when the transformer is running without changing the internal insulation structure of the transformer,the fiber optic is fixed on the coil by using epoxy resin and insulating paper specially designed for transformers,and the 35 kV winding model with built-in vectorized fiber optic sensor is successfully tested,and the deformation monitoring of the windings is completed by applying six groups of axial and radial deformation of different degrees by human beings,and by using the time-domain reflection technology of Brillouin light.The deformation of the winding is monitored by Brillouin's optical time-domain reflection technique.The experimental results show that the mean value of the monitoring error of axial deformation is 9.63%,and the monitoring error of radial deformation is 5.18%,which is mainly caused by the uneven force when applying the force as well as the equipment acquisition error,and the error tends to be stabilized after the deformation is increased,and the experimental results effectively verified the feasibility of the vectorized monitoring model of the winding deformation,and realized the direction of the deformation of the transformer and the degree of demodulation of the deformation,which further improves the monitoring accuracy of the winding deformation.The monitoring accuracy of winding deformation is further improved.关键词
绕组形变/矢量化监测/光纤传感/简支绕组横梁Key words
Winding deformation/vectorized monitoring/optical fiber sensing/simply supported winding beam分类
信息技术与安全科学引用本文复制引用
范晓舟,袁洁平,薛峰,王湘女,律方成,耿江海..基于简支绕组横梁算法的变压器绕组形变矢量化监测模型[J].电工技术学报,2025,40(5):1639-1651,13.基金项目
中央高校基本科研业务费专项资金(2023MS106)和南方电网公司科技项目(031900KK52220012)资助. （2023MS106）
