基于高阶累积量的改进Prony低频振荡辨识方法

摘要

广域测量系统的不断发展和进步,为基于Prony方法实现电力系统低频振荡的在线辨识提供了可能性.针对传统Prony算法在辨识广域测量系统监测信号过程中抗噪声性能不足的问题,本文提出一种基于高阶累积量的改进Prony算法用于低频振荡的在线辨识.该方法利用高阶累积量技术具备的良好抗噪特性,能够有效抑制广域测量系统在采集和传输过程中产生的白色噪声和有色噪声.最后,利用典型正弦信号、新英格兰10机39节点算例系统的仿真结果和实际电网测量数据验证了所提方法的正确性和有效性.%With the continuous development and improvement of wide area measurement system(WAMS),it is possi?ble to realize online detection of low frequency oscillation in power system using Prony algorithm. To overcome the poor anti-noise performance of the traditional Prony algorithm when it is used in the detection of WAMS monitoring signals , an improved Prony algorithm based on higher-order cumulant(HOC)is proposed in this paper. It is more effective in re?straining the white and colored noises generated in the acquisition and transmission process of WAMS,since HOC has better anti-noise performance. Finally,simulations on typical sinusoidal signals,New England 10-generator 39-node system and the data from an actual power grid verify the correctness and effectiveness of the proposed method.