基于界面电荷极化特性的变压器油纸绝缘寿命预测方法研究

准确预测变压器油纸绝缘寿命对于保障电力系统安全稳定运行具有重要意义。根据微观介电响应机制剖析界面极化与油纸绝缘老化的内在机理联系，并基于混合极化模型——一种在扩展德拜模型基础上引入界面极化支路的等效电路拓扑，探究界面电荷极化特性在油纸绝缘寿命预测中的应用。首先，分析温升过程中频谱特性变化背后蕴含的普适弛豫极化规律，基于混合极化模型重构回复电压极化谱。结合温度对主时间常数和模型参数的影响规律，提出一种单一温度测试下的油纸绝缘活化能计算方法，通过界面极化电容构建出频温平移因子。其次，剖析介质损耗因数随老化加深的变化规律，探究损耗峰偏移现象与界面极化弛豫时间的内在联系，凝练全新表征油纸绝缘聚合度含量的频域特征参量——界面极化支路极点Ph2。最后，联合界面极化支路极点Ph2、频温平移因子和损失累积动力学方程构建油纸绝缘寿命预测模型，并将不同运行年限变压器现场实测数据代入模型验证实效性，为油纸绝缘寿命预测提供了一种新思路。

A life prediction method of transformer oil-paper insulation based on interfacial charge polarization characteristics

Accurate lifetime prediction of transformer oil-paper insulation is of great significance in ensuring the safe and stable operation of a power system. In this paper, the internal mechanism relationship between interfacial polarization and aging oil-paper insulation is analyzed based on a microscopic dielectric response mechanism. Based on a mixed polarization model, an equivalent circuit topology with an interfacial polarization branch is introduced on the basis of an extended Debye model, and the application of interfacial charge polarization characteristics in oil-paper insulation life prediction is explored. First, the universal relaxation polarization law behind the spectrum changes in the process of temperature rise is analyzed, and the recovery voltage polarization spectrum is reconstructed based on the mixed polarization model. Combined with the influence law of temperature on the principal time constant and model parameters, a calculation method for the activation energy of oil-paper insulation in a single temperature test is proposed. The frequency-temperature translation factor is constructed from the interface polarization capacitance. Secondly, the change rule of dielectric loss factor with the deepening of aging is analyzed, the internal relationship between loss peak shift and relaxation time of interfacial polarization is explored, and a new characteristic parameter in frequency domain, the interfacial polarization branch poleis refined to characterize the polymerization degree content of oil-paper insulation. Finally, an oil-paper insulation life prediction model is constructed by combining interface polarization branch pole, frequency-temperature translation factor and a loss accumulation dynamic equation, and the measured data of a transformer of different operating years are substituted into the model to verify its effectiveness. This provides a new idea for oil-paper insulation life prediction.