基于模型识别的配电网谐波源定位方法

Harmonic Source Location Method for Distribution Networks Based on Model Recognition

数量日益增加的谐波源将威胁电力系统的安全稳定运行,谐波源的定位与谐波治理意义重大。该文首先分析配电网中各类设备的输出特征,指出非谐波源具有线性特性,等效为电阻–电感串联电路模型,而谐波源具有非线性特性,等效为非线性电路模型。利用谐波源与非谐波源等效模型间明显的非线性与线性差异,可以有效地区分谐波源与非谐波源。由此提出基于模型识别思想的谐波源定位方法,计算识别设备量测电气量与线性模型的符合程度,二者符合程度高即为非谐波源,否则为谐波源。同时,定义模型识别结果的波动程度来定量描述谐波源的非线性度。在IEEE 13节点系统进行性能验证,结果表明,该文提出的谐波源定位方法可准确定位不同种类的谐波源和组合谐波源,定义的非线性度指标能有效反映谐波源的谐波贡献度。最后对比分析提出的模型识别法与现有的谐波状态估计法、功率方向法的性能,验证的结果表明模型识别法在定位准确度和数据需求等方面均具有明显优势。

Numerous harmonic sources threaten the safe and stable operation of the power system. Therefore, it is of great significance to locate and control the harmonic sources. In this paper, the output characteristics for different types of devices in the distribution network were analyzed firstly. It could be concluded that the non-harmonic sources have linear characteristics and could be equivalent to resistance inductance series circuits, while harmonic sources have nonlinear characteristics and could only be equivalent to nonlinear circuits. Utilizing the obvious nonlinearity and linearity difference between the equivalent models of the harmonic source and the non-harmonic source, the harmonic source could be effectively distinguished from the non-harmonic source. Therefore, the model recognition-based harmonic source location method was proposed. To calculate the consistent degree between the measured electrical variables of the device and the linear model, and a high consistent degree between the two is a non-harmonic source, otherwise it is a harmonic source. Moreover, the nonlinear degree of harmonic source was quantitatively described by defining the fluctuation degree of model recognition results. The proposed method was verified in IEEE 13 bus test system and the test results have indicated that this method could accurately locate different types of harmonic sources and combined harmonic sources, and the defined nonlinearity index could effectively reflect the harmonic contribution of the harmonic source. Finally, compared with the harmonic state estimation-based method and power direction-based method, the proposed method has obvious advantages in positioning accuracy and data requirements.