论文部分内容阅读
传统电力系统次同步振荡的辨识方法存在对噪声敏感、辨识精度不高的局限性。为此,提出了一种基于数学形态学自回归移动平均(MM-ARMA)算法的辨识方法,实现了在有噪声干扰下对次同步振荡模态的准确辨识。该方法利用形态滤波器可以有效抑制噪声的特性对次同步振荡信号进行消噪处理,保留信号的主要特征信息;对消噪后的信号建立基于加权递推最小二乘法参数估计的ARMA模型,根据估计的模型参数计算次同步振荡模态参数,完成次同步振荡模态辨识。与传统的Prony算法和自回归移动平均(ARMA)算法辨识结果进行的对比分析结果表明,所提次同步振荡模态辨识方法能快速、准确地辨识出模态参数,且具有较强的抗噪能力。
The identification method of subsynchronous oscillation in traditional power system is sensitive to noise and its identification accuracy is not high. Therefore, an identification method based on the mathematical morphology autoregressive moving average (MM-ARMA) algorithm is proposed, which realizes the accurate identification of subsynchronous oscillation modes under noisy interference. This method uses the morphological filter to effectively suppress the characteristics of noise to denoise the subsynchronous oscillation signal and preserve the main characteristic information of the signal. An ARMA model based on weighted recursive least square method is established for the de-noised signal. According to the ARMA model, The estimated model parameters calculate the sub-synchronous oscillation modal parameters to complete subsynchronous oscillation mode identification. Compared with traditional Prony algorithm and autoregressive moving average (ARMA) algorithm, the results show that the proposed sub-synchronous oscillation mode identification method can quickly and accurately identify the modal parameters and has strong anti-noise ability.