原子稀疏分解算法在电力系统扰动信号分析中的应用

Application of atomic sparse decomposition to power systems disturbance analysis

针对目前电力系统扰动信号分析中的一些不足,提出了一种基于原子分解的扰动信号分析方法。该方法根据电力系统信号特点,采用衰减正弦量模型构建原子库。以原子稀疏分解的基本理论为指导,应用匹配追踪算法进行迭代求解,得到相关原子参量。将所述方法应用于电力系统扰动信号分析中,能够自适应地从过完备原子库中建立信号的稀疏解析表示,克服了固定基信号分解模式的缺陷。典型算例仿真结果证明了算法的有效性和优越性。

An atomic decomposition based method is introduced to solve power system disturbance signal analysis problems.Considering the features of power system signals,the damped sinusoid atom is used to construct atom dictionary.The principle of sparse decomposition is referenced and the matching pursuit algorithm is applied to obtain the relevant parameters.Atomic decomposition is capable of adaptively representing signals in an over-complete dictionary of atoms with very sparse components rather than with fixed and complete basis as in conventional representations.Examples of applying this method to several typical power system disturbance analyses exhibit the validity and the great advantage. This work is supported by National Natural Science Foundation of China （No.50877069）.