基于支持向量回归的直流受端电网动态无功需求在线评估

暂态电压稳定性已经成为直流受端电网安全的重要威胁之一,实时在线的动态无功源管理和控制是应对这一挑战的有效手段。因此,有必要研究直流受端电网的动态无功需求在线评估技术,量化评估动态无功设备对电网暂态电压稳定性的支撑作用。以往,主要依赖时域仿真法对大电网的暂态电压稳定性进行分析,由于计算耗时较长等原因导致其在线应用受限。针对这一问题,提出了基于支持向量回归(SVR)的直流受端电网动态无功需求在线评估方法。首先,建立了能够量化评估电网暂态电压稳定性的STVSI-SVR模型,然后提出了以该模型为基础的动态无功储备评估算法。相比于传统的基于时域仿真的分析方法,该算法能够快速地计算出合理的结果,满足在线应用的需求。最后在我国某实际受端电网模型上测试了算法的有效性。

Online evaluation of dynamic reactive power demand of DC receiving power grid based on SVR

Transient voltage stability has become one of the important threats to the safety of DC receiving power grids. Real-time online dynamic reactive power source management and control is an effective means to cope with this challenge. Therefore, it is necessary to study the dynamic evaluation of dynamic reactive power demand of DC receiving power grid, and quantitatively evaluate the supporting role of dynamic reactive power equipment on grid transient voltage stability. In the past, it mainly relied on the time domain simulation method to analyze the transient voltage stability analysis of large power grids. It is difficult to realize online application due to the long time consumption of calculation. In response to this problem, this paper proposes an online evaluation method for dynamic reactive power demand of DC receiving power grid based on Support Vector Regression (SVR). Firstly, the STVSI-SVR model which can quantitatively evaluate the transient voltage stability of the power grid is established, and then the dynamic reactive reserve evaluation algorithm based on this model is proposed. Compared with the traditional time domain simulation based analysis method, the algorithm can quickly calculate reasonable results and meet the needs of online applications. At the end of the paper, the effectiveness of the algorithm is tested on a practical receiver network model in China. This work is supported by Science and Technology Project of China Southern Power Grid Ltd. (No. GDKJ00000060).