A power grid partitioning optimization method based on fractal theory

It is important to identify characteristics of power grid structures in many applications. In a power system, large grids are divided into areas to meet planning and control requirements. Topological structures are considered in partitioning methods to guarantee the reliability and extendibility of power grids. However, the representation of structures is complex. To apply topological structures efficiently, this paper presents a new method for quantizing topological structures of power grids based on fractal theory. The fractal property of power grids is proved, and the fractal dimension of power grids is measured by a box-counting algorithm. A fractal dimension-based prediction method of short-circuit current is proposed to identify problem regions quickly. Then a power-grid-partitioning optimization method based on the quantization results of power grids is introduced for power grids. Finally, the IEEE 118-bus test system and an actual power grid in China are employed to illustrate the validity of the proposed method.