Data-Driven Short-Term Voltage Security Constrained Reactive Power Optimization Method

The goal of short-term voltage security (STVS) constrained reactive power optimization is to improve the STVS of the power grid by regulating the operating state of reactive power resources. However, the STVS constraints are high in dimension and contain differential-algebraic equations, so the optimization problem is difficult to solve. To this end, this paper establishes a machine learning model for STVS analysis and proposes a data-driven STVS constrained reactive power optimization method. A simplification for STVS constraints is proposed to reduce the difficulty of solving the optimization problem. The proposed model is tested based on the IEEE-39 model. The results show that: the proposed method can effectively improve the STVS of the power grid, and the calculation speed is fast.