Access to scattered solar panels has transformed the originally passive distribution network of East China’s Shandong Province into an active one. At noon, power generation mounts while consumption plummets, resulting in a reverse power flow that frequently causes overloads on distribution transformers.
Conventional manual methods of dealing with this issue are time-consuming and imprecise. In response, the Shandong branch of State Grid Corporation of China (State Grid) has launched a pilot smart inspection and control scenario in the city of Dezhou in Shandong that is tailored for the high-proportion connection with distributed photovoltaic (PV) resources.
Supported by two new systems that can respectively manage power load and collect information on power consumption, the scenario has deployed four intelligent agents for PV anomaly identification, reverse overload cause analysis and regulation strategy formulation and optimization. It has established a closed-loop management model featuring cloud-based intelligent analysis and on-site automatic inspection and control.
The scenario has boosted the identification rate of PV equipment anomalies to over 95 percent and regulation accuracy to over 97 percent, and it has shortened the regulation time to under 10 minutes.
With the rapid growth of new-type power loads and user-side resources, conventional mechanisms fail to deal with the location, volume and time of adjustable power resources. To solve this problem, State Grid’s Shanghai branch has built a hierarchical, collaborative mechanism that has cloud-based large models, lightweight small models and terminal-based intelligent agents, and has adopted the Smart Energy Management Master System as the technical core to support the intelligent aggregation and precise regulation of virtual power plants (VPPs).
The system can first distribute regulation tasks to various operational entities. After they respond, the system tracks real-time load changes and optimizes dispatching strategies. Then the system automatically verifies regulation effects and provides auxiliary support for revenue settlement.
Up to now, the scenario has incorporated multiple resources, such as user-side energy storage, and has been applied in various fields, such as response to summer peak load, new energy consumption and cross-provincial transfer of computing power driven by spot electricity prices.
Since the winter peak period of power consumption in 2024, the scenario has supported 95 rounds of VPP responses in Shanghai, cumulatively delivered over 7.2 million kilowatt-hours of power and covered 61 VPP operators and more than 20,300 end users.
(Executive editor: Zuo Shihan)
