Improved VVO Algorithm Based on Multi-Node Voltage Sensitivity Control in Distribution Networks

The increasing prevalence of distributed photovoltaic (PV) and energy storage systems in distribution networks, coupled with the rising load fluctuations, has led to an elevated risk of voltage violations at network nodes. This study investigates methods to enhance the controllability of node voltages based on the Voltage and Var Optimization (VVO) algorithm. While theoretically VVO can yield favorable outcomes in voltage limit regulation, its control process heavily relies on the objective function and optimization algorithm. Due to its intricate nonlinearity and multi-objective optimization, maintaining accuracy in voltage regulation on a global scale is not a trivial task. Therefore, a sensitivity-based improved VVO algorithm is proposed to compute real-time actions for energy storage and Line Voltage Regulator (LVR), aiming for more precise voltage control. The VVO algorithm is an approach for voltage and reactive power optimization in power systems, adjusting the reference values of reactive power in the grid to achieve optimal distribution of voltage and reactive power. However, it may converge to local minima and fail to reach a global optimum. To address this issue, an enhanced VVO algorithm is introduced to optimize the actuation of control devices. Initially, the consistency between topology modeling of power flow and actual consumption data is verified to ensure the accuracy of line parameters. Subsequently, a control strategy involving the participation of sensitivity for all node voltages is proposed to prevent device actions from getting trapped in local optima. Through this approach, a novel voltage control method named Improved VVO (IVVO) is constructed. The results demonstrate that IVVO outperforms VVO, exhibiting the fastest algorithm convergence speed, minimal reactive and active control quantities for energy storage, and the lowest line active and reactive losses. In comparison with common voltage control models such as Clustered Real-Time Adaptive Model-Based Control (MBC) and Second-Order Oscillating Particle Swarm Optimization with Constriction Factor (SOOPSO), the proposed IVVO demonstrates superior overall performance in terms of convergence speed, highest and lowest voltage accuracy, and energy storage active and reactive variations, with values of 11 iterations, 0.001%, 0.003%, 5.409 kVar, and 4.574 kW, respectively.