Co-Simulation-Based Evaluation of Control Mechanisms for Active Low-Voltage Network Scheduling
This paper proposes a co-simulation model for the operation of an active distribution network scheduling, incorporating three objective functions: voltage deviation minimization, losses minimization and the voltage unbalance factor minimization. The developed model accounts for the available operational control mechanisms in an active low-voltage distribution network, including conventional transformer tap changer and the prosumers’ inverter capabilities with emphasis on active power curtailment and reactive power control. The proposed model is formulated within a co-simulation environment, which enables the application of computational intelligence methods to obtain near-optimal solutions. In this paper, particle swarm optimization is used as a computational intelligence method. The evaluation of the proposed model is assessed on a real-world low-voltage distribution network. To analyse the impact of the various control mechanisms across all three objective functions, 45 case studies are carried out. The impact of each control mechanism on achieving a particular objective function is analyzed, assuming the remaining two variables defining objective functions are treated as a state. The results indicate that low-voltage distribution network parameters have a strong impact on control mechanism effectiveness, independent of the selected objective function, and notably favor the active power curtailment. Under the proposed control, energy losses decreased from 164.98 kWh to 6.09–14.33 kWh (92.70-96.31%) for different objective functions. Voltage limits are maintained within a 0.92–1.09 p.u. band (compared to 0.85–1.19 p.u. in the reference case), and the VUF is reduced from 2.41% to 0.46%–1.52%. Considering existing regulations and technical limits on reactive power, reactive power control is insufficient to handle significant deviations. Finally, voltage deviation minimization and voltage unbalance factor minimization each have notable limitations in their operation scheduling performance.