时变温度下油纸绝缘频域介电谱曲线校正方法研究

Study on the Frequency Domain Spectroscopy Curves Correction Method of Oil-Paper Insulation at Time-Varying Temperature

频域介电谱(FDS)技术被广泛应用于油纸绝缘电力设备状态诊断中。但FDS在低频段测试时间较长,测试时设备往往处于动态降温过程,所获得的FDS曲线与评估数据库差别较大,使用时变温度FDS曲线进行绝缘状态评估的结果存在一定误差。为将时变温度条件下的FDS曲线校正到恒定温度,该文通过仿真获得套管温度场分布,采用粒子群优化(PSO)算法反演出不同温度下介电弛豫Havriliak-Negami(H-N)模型的特征参数。在此基础上,基于最小二乘法(LS)计算出时变温度条件下每个频点的等效温度,由此实现了曲线校正。结果表明:校正后的曲线与参考温度曲线具有较好的一致性,有助于电力设备绝缘状态的精确评估。

In recent years,frequency domain spectroscopy(FDS)has been widely used in the insulation status diagnosis of oil-paper bushing.However,at present,FDS testing can only be realized at steady temperature.The environment under actual working conditions is complex.Especially affected by external factors such as region and climate,the equipment is often in a dynamic cooling process when it is out of service for maintenance.In addition,due to the short“time window”of on-site maintenance,the time for FDS testing is limited.Therefore,it is impossible to conduct dielectric response assessment after the bushing temperature is stabilized,and the obtained FDS curves are quite different from the constant temperature curves.In order to solve these problems,this paper obtains the equivalent temperature of each frequency point of the FDS curves under the condition of time-varying temperature based on Havriliak-Negami(H-N)model,so as to realize FDS curve correction and improve the accuracy of insulation state evaluation.Firstly,the heat dissipation simulation model was constructed with full consideration of the real structure of the casing.On this basis,the conduction heat dissipation inside the bushing capacitor core,the external convection heat dissipation of the vertical wall and horizontal wall,and the radiation heat dissipation of the outer sheath surface to the environment were studied,and the temperature distribution characteristics of the bushing under different heat dissipation times were obtained.It is found that with the increase of heat dissipation time,the insulation temperature inside the bushing does not decrease at a constant rate,but shows an exponential downward trend,and the heat dissipation rate changes from fast to slow.In addition,the experiment shows that the error between the temperature curve of bushing oil temperature changing with time and the simulation curve is less than 6%,which verifies the accuracy of simulation.Furthermore,the FDS curves of several groups at constant temperature were obtained through testing.H-N relaxation function can accurately describe the relaxation process including interface polarization.Therefore,the nonlinear relationship between the characteristic parameters of the H-N dielectric relaxation model of the constant temperature FDS curves and temperature was inversed by using the particle swarm optimization(PSO)algorithm.Since the equivalent temperature cannot be expressed as an explicit expression,and the known data are more than the number of parameters to be solved,this paper uses the least square method(LS)to calculate the equivalent temperature of each frequency point in the FDS curves under the condition of time-varying temperature.Finally,the Arrhenius formula was used to achieve accurate correction of frequency point by point data under transient temperature.With 30℃,60℃and 90℃as the reference temperature,the time-varying temperature curves under three different conditions were corrected.The results show that compared with the test results under 30℃,60℃and 90℃,the test results under time-varying temperature conditions are distorted in the low frequency part,and the curve shows a downward shift trend.The calibration results under three conditions are basically consistent with the test results under 30℃,60℃and 90℃,and the curve overlap well,which verifies the accuracy of the calibration model.Therefore,it can be concluded that the proposed correction method can effectively correct the test results under time-varying temperature conditions,which is helpful to the accurate evaluation of the insulation status of power equipment.