基于改进Prony算法的新能源配网谐波和振荡复合扰动参数提取

Extraction of Harmonic and Oscillation Composite Disturbance Parameters of New Energy Distribution Network Based on Improved Prony Algorithm

新能源接入含大量非线性负载的配电网,使其背景谐波更为复杂,甚至导致谐波谐振,严重影响配电网稳定运行。准确提取宽频振荡扰动及背景谐波特征参数是谐波振荡识别与治理的关键技术。针对传统的Prony算法提取扰动信息精度易受噪声与有效秩阶数的影响,提出奇异值分解法降噪法改善宽频振荡扰动信号质量。基于二阶导数理论的矩阵奇异值排列方法,选取合适模型阶数,提高背景谐波参数的辨识精度。仿真结果验证了所提算法的有效性。

The access of new energy to a distribution network with a large number of nonlinear loads makes the corre⁃sponding background harmonics more complex and even leads to harmonic resonance,which seriously affects the stable operation of the distribution network.Therefore,the accurate extraction of broadband oscillation disturbances and back⁃ground harmonic characteristic parameters is a key technology for harmonic oscillation identification and management.Since the accuracy of the traditional Prony algorithm when it is used to extract the disturbance information is easily af⁃fected by noise and effective rank order,the singular value decomposition noise reduction method is proposed to im⁃prove the broadband oscillation disturbance signal quality.Based on the second derivative method,the matrix singular value arrangement method is used to select an appropriate model order to improve the identification accuracy of back⁃ground harmonic parameters.Simulation results show that the proposed algorithm is effective.