基于EMD去噪的Prony算法在电网事故分析中的应用

基于DSA系统,江苏省电力调度控制中心开展了电网事故反演分析。在仿真曲线与PMU录波的比对分析过程中,发现PMU信号在测量、传输过程中引入的随机噪声可能严重干扰Prony分析结果的精确性。文中提出应用基于噪声统计特性的EMD算法作为Prony分析的前置滤波环节,通过随机重排循环,对EMD分解得到的噪声分量进行抑制,大幅降低噪声能量,提高重构信号的信噪比,从而改善Prony算法的分析精度。通过测试信号与江苏电网实际故障后分析算例,验证了算法的有效性与实用性。

The Application of EMD-based Prony Algorithm in Power Grid Fault Analysis

With the DSA system, Jiangsu Electric Power Dispatching and Control Center implemented an after-fault analysis.Compared PMU-recorded data with simulation curve, it is discovered that random noise in the PMU data, introduced during the process of measuring, transmission and conversion etc., may significantly affect the precision of Prony analysis. This paper proposes a noise characteristic based EMD which can be used as a pre-filter algorithm in Prony analysis. This algorithm reduces the power of noise through random sort iterations, and thus increases the signal-noise-ratio of the reconstructed signal.Therefore, the noise in the signal can be filtered effectively within several loops. The comparison between a test signal and a real fault in of Jiangsu power grid proved the effectiveness and practicability of the proposed algorithm.