Power system transient stability margin assessment using steady-state information

Transient stability margin assessment provides power grid dispatchers with important criteria for decision-making and preventive control. The computational efficiency of online transient stability margin assessment can be largely improved with the use of steady-state information. In this paper, we propose a novel methodology for online stability margin assessment. A multi-swing based strategy that overcomes the false identification issues of previous methods is firstly illustrated. Based on this improved strategy, a composite neural network is constructed and trained offline to establish the mapping relationship between steady-state characteristics and generator transient stability indices. Finally, the trained neural network is utilized for online transient stability margin assessment. The major advantage of the proposed methodology is that we only use steady-state information in the online assessment process therefore system transient stability margin under worst disturbances can be fast and accurately predicted, providing system dispatchers with adequate time for preventive control. The validity of the proposed methodology is demonstrated in the IEEE 39-bus system.