Edge Computing Architecture for Monitoring Distribution Networks with Distributed Photovoltaic Integration

Abstract

The increasing penetration of Distributed Energy Resources, particularly photovoltaic (PV) systems, is imposing critical requirements on low-latency Volt-Var Control (VVC) solutions in the distribution grid. Traditional Centralized Control Architectures (TCCs) are obsolete due to the transmission and calculation delays encountered when solving complex global optimization problems. This paper proposes and analyzes the Cloud-Edge Computing-Cooperative VVC Architecture (CEC-VVC) for the IEEE 33-bus power system. CEC-VVC achieves an optimal balance by utilizing Edge Servers for local optimization within the sub-zone via the MISOCP and by implementing real-time reactive power control of PV inverters using the ANN algorithm. Simultaneously, the system performance is analyzed quantitatively using the data splitting ratio. In-depth analysis shows that the CEC-VVC architecture has reduced the system's Total Response Time from 609ms (TCC) down to a minimum of 121.8ms, achieving an 80% improvement and meeting the real-time response criteria. At the same time, this data splitting ratio reduces the data capacity stored in the Cloud by 2.5 to 10 times compared to TCC. More importantly, the sensitivity analysis proves the robustness of the architecture. The optimal operating point is maintained even when the data transmission speed changes significantly. These results confirm that CEC-VVC provides a flexible balance solution between response speed and resource efficiency, making it an important proposal for stable and reliable operation of the smart grid in the future.