SE-CNN based emergency control coordination strategy against voltage instability in multi-infeed hybrid AC/DC systems

For receiving-end power systems with multi-infeed HVDCs and large-scale renewable energy generation, the challenges of weak reactive power support and limited anti-interference capacity pose a threat to the power system voltage stability. To enhance the voltage stability under large disturbances, this paper proposes a data-driven method for coordinated emergency control strategy against voltage instability in multi-infeed hybrid AC/DC systems based on the squeeze and excitation (SE) module and convolutional neural network (CNN). Firstly, the SE module and CNN are integrated to intuitively capture correlations among various feature channels and extract key features related to voltage stability levels. Secondly, a voltage stability evaluation method is proposed based on the SE-CNN model, utilizing a dual channel structure to unveil the quantitative relationship between key response characteristics and voltage stability margin. Finally, regarding the control measures containing load shedding and DC modulation, the emergency control sensitivity index is proposed to assess the stability margin improvement effect of various control objects, and the optimal emergency control strategy is formulated to coordinate multiple voltage stability emergency control measures. Case studies were conducted on a typical China local region receiving-end power grid test system with voltage instability problems, and the simulation results verified the effectiveness of the proposed method.