Key points are not available for this paper at this time.
The application of effective voltage control methods is currently a critical issue in active electrical distribution networks (DNs). This paper proposes a novel hierarchical system with mutual support between medium-voltage (MV) and low-voltage (LV) DNs for voltage optimization, based on a holonic architecture. The application of multi-agent systems (MAS) in active DNs can facilitate communication between distributed generation sources and various other agents, such as capacitor banks (CBs) and tap changers. This study is conducted to optimally size distributed generations (DGs) into an MV and LV DN to minimize power losses and voltage deviation. For considering the actual power system scenarios, a penalty factor is also considered. The proposed method is implemented using the genetic algorithm. The effectiveness of the proposed architecture was investigated in a case study network comprising an MV network connected to three LV DNs. The proposed architecture was compared with three baseline architectures (without distributed generation, centralized control, and holonic architecture) using a set of evaluation indices. Simulations were conducted using the MATPOWER 8.0 toolbox in MATLAB. The obtained results show that, simultaneously considering both LV and MV networks and employing their respective agents, active power losses and voltage deviations are reduced compared to the centralized and holonic architectures. It has shown the effectiveness of the proposed method. The reduction in the sum of voltage deviation at peak load and losses during the hours when it was highest is more significant. Also, the maximum voltage value has been reduced from 1.049 pu to 1.036 pu.
Ardakan et al. (Sat,) studied this question.