Candidate bus identification for voltage stability enhancement of wind-penetrated power system based on spectral learning technique

Dynamic reactive power sources can efficiently address the voltage stability problem of a wind-penetrated power system and the possibility of cascading failures, which are caused by the decreasing inertia and the increasing complexity of the system dynamics. However, their applications are limited by the high investment cost. Identifying appropriate buses for the deployment can improve the economy and efficiency of reactive power source configuration and reduce the complexity of deployment models. Here, a new metric is proposed to guide the selection of candidate buses, based on the improved spectral learning technique and the quantitative assessment of the short-term voltage stability. Specifically, a new short-term voltage stability metric is developed to assess the dynamic voltage responses of different stages after a contingency. Then, an improved spectral learning algorithm with objective priorities assigned to different buses is used for the bus selection, aiming to identify the most influential buses, in terms of short-term voltage stability and propagation potentials. A two-dimensional decision-making methodology is proposed, considering both the capacity sensitivity and the bus's structural characteristics. The effectiveness of the proposed methodology is validated on a New England 39-bus system using an electromechanical transient model.