中文摘要：

对工作于非线性负荷电网下的电子式互感器进行在线校验,数据处理算法的精确度是保证在线校验可靠性的关键。为提高在线校验精度,有效抑制频率波动、间谐波和高次谐波的影响,提出一种基于离散Fourier变换（DFT）算法数据预处理及主频谱修正的改进算法。在离散Fourier变换前对采样数据进行等长度截断叠加,实现相位补偿;构建二阶Hanning自卷积窗函数,对截断采样序列进行离散Fourier变换并提取基频分量,通过主频谱校正得修正后的幅值和相位。采用虚拟仪器开发环境Labview进行仿真,计算了不同叠加次数对改进算法输出值误差的影响,并对几种不同的数据处理算法进行了对比。研究结果表明：叠加次数α=3时,改进算法输出值的误差最小;当电网频率出现±0.5 Hz的频率波动时,改进算法的精确度更高;性能测试结果表明,比差变化〈0.01%,角差变化〈0.3′,误差分布均匀且无跳变,准确度可达到0.05级,满足对0.2S级电子式互感器的校验要求。研究结果证明了所提算法的有效性。

英文摘要：

For the purposes of on-line calibration on electronic transformer that works in nonlinear complex grid environment, the accuracy of the data processing algorithm is critical to ensuring the reliability of the on-line calibration. In order to improve the accuracy of on-line calibration, and to effectively inhibit the effects of frequency fluctuations, harmonic wave, and higher harmonics, we put forward the data preprocessing and the main spectrum modification improved algorithm based on discrete Fourier transform（DFT） algorithm. For the purpose of phase compensation, the isometrictruncated sample data before DFT were also added. By constructing the second-order Hanning self-convolution window function, we could truncate the sampling sequence of discrete Fourier transform and extract the fundamental frequency component. Then the corrected amplitude and phase through the main spectrum correction were obtained. The virtual instrument development environment labview was used to simulate the influences of different stacking times on the output error of the improved algorithm, and several different data processing algorithms were compared. The results show that the error of the improved algorithm is the smallest when the value of superposition time α is 3, and the accuracy of the improved algorithm is higher when the frequency fluctuates by ±0.5 Hz. Performance test results show that the ratio deviation change is less than 0.01% and the angle deviation change is less than 0.3′. The deviation is evenly distributed and no transition occurs. The accuracy of on-line calibration system attains to 0.05 levels, which meets the calibration requirements of the 0.2 s levels electronic transformer. The results show that the proposed algorithm is effective.