基于混合放电模型的变压器内部放电联合检测方法研究

Study on joint detection method of transformer internal discharge based on hybrid discharge model

随着电力设备运行检测及监测技术的发展,变压器作为关键主设备其设备内部绝缘缺陷的检测成为研究的重点和方向,但是由于变压器内部结构的复杂性、不同绝缘缺陷复现的难度、局部放电信号干扰信号难以辨识等问题,造成目前对于变压器内部不同绝缘缺陷检测、分析及辨识的水平仍然不高。为了探索有效的检测方法,文中开展了针对性的研究工作,基于变压器故障仿真模型,对变压器内部绝缘缺陷的模拟技术开展研究,实现变压器内部缺陷的真实环境下的准确复现,设计了两类变压器故障混合模型,通过基于HFCT、UHF、AE、光学四种传感器的复合信号采集系统,进行变压器局放过程中基于不同信号原理的信号特性研究。同时基于采集信号对不同放电信号特征进行关联性分析,实现多信号特征的综合分析,进一步提高局放信号辨识真确度,研究表明文中所设置的局放复合信号采集系统能准确测量变压器中的局部放电信号,并反映其各类信号的信号特征;通过对比单一局放模型和混合局放模型下局部放电发生时的信号特征、能量分布特征等,为今后的变压器混合局部放电模式识别研究奠定了基础。

As the development of power equipment operation detection and monitoring technology,the transformer,as a key master device,the detection of insulation defects inside the equipment has become the focus and direction of the research.However,due to the complexity of the internal structure of the transformer,the difficulty of reproduction of different insulation defects,the difficulty of identifying the local discharge signal interference signal and other issues,the level of the transformer internal insulation defects detection,analysis and identification is still not high at present.In order to explore the effective detection method,this paper has carried out targeted research work.Based on the simulation mode of transformer failure,research on simulation technology of transformer internal insulation defect has been carried out,and two kinds of transformer fault hybrid models are designed to realize the accurate reproduction in the real environment of transformer internal defects.Through the composite signal acquisition system based on HFCT,UHF,AE,optical four kinds of sensors,research on signal characteristics based on different signal principles in transformer partial discharge has been carried out.Meanwhile,the correlation analysis of different discharge signal characteristics based on the acquisition signal is carried out to realize the comprehensive analysis of multi-signal characteristics and further improved the accuracy of PD signal identification.The research shows that the PD signal acquisition system can accurately measure the partial discharge signal in the transformer and reflect the signal characteristics of all kinds of signals.By comparing the signal characteristics and the energy distribution characteristics of the local discharge under the single PD model and the hybrid PD model,it provides a basis for the future study of the hybrid partial discharge pattern recognition of the transformer.