基于广域发电机状态信息的电力系统暂态不稳定性实时预测

Real-time prediction of transient instability based on wide-area information of generator state

基于两机群等值方法,将相平面轨迹几何特征判据推广应用于多机系统暂态不稳定性实时预测。为实现该不稳定性预测方案在可靠性和快速性的折衷,着重研究了失稳模式实时筛选和基于预测轨迹的不稳定性判别2项关键性技术。失稳模式实时筛选技术利用角度间隙作为指标选择主导模式搜索方向,并采用不稳定性条件作为搜索终止判据,从而使得搜索中所得主导模式随着轨迹发展逐渐逼近真实失稳模式;基于预测轨迹的不稳定性判别技术将只在参数时变性恶化系统稳定性时进行等值轨迹的预测,并基于该预测轨迹的几何特征判断系统是否稳定。IEEE 145节点系统和某实际电网系统的仿真计算验证了所提预测方案的有效性。

Based on the dual-cluster equivalence method, the geometric characteristic criterion of phase plane trajectory is applied to the real-time prediction of transient instability for multi-machine system. For the compromise between speediness and reliability of the prediction scheme,two key technologies are studied： the real- time filtration of instability mode and the instability detection based on predicted trajectory. The angle gap is used as an index to choose the dominant mode search direction and the instability conditions to stop the dominant mode search, which makes the dominant mode approaching the real instability mode step by step along with the trajectory evolution. The equivalent trajectory prediction is carried out only when the parameter variation may weaken the system stability,and the geometric characteristics of the predicted trajectory are used to determine if the system is stable. Numerical simulations on IEEE 145-bus test system and some provincial power grids show the effectiveness of proposed prediction scheme.