In recent years, renewable energy generation such as wind power has been widely applied in distribution networks. However, its output randomness leads to inaccurate power flow optimization results. To address the reactive power optimization problem of distribution networks under such uncertain operating conditions, a dynamic reactive power optimization method is proposed by incorporating the real-time regulation characteristics and operational constraints of the power system automatic voltage control (AVC) system. A day is discretized into 24 time intervals, and a dynamic probabilistic reactive power optimization model for distribution networks considering multiple uncertain factors is constructed, with the objective of minimizing the sum of expected values of the system's active power loss across all time intervals. For the solution of the established model, an improved grey wolf optimizer is adopted to overcome the premature convergence defect of traditional optimization algorithms, so as to derive the optimal solutions of deterministic control variables and the expected value of system power loss. Simulation calculations on the IEEE 33-bus system verify the feasibility of the proposed model and the effectiveness of the algorithm. • Proposes an AVC-based dynamic reactive power optimization model for wind farm grids to address new energy volatility. • Enhances voltage regulation and reactive power balance via coordinated multiple sources using an improved control strategy. • Validated on IEEE 33-node system: the model reduces grid loss and stabilizes voltage deviation within safe limits.
Yang et al. (Mon,) studied this question.