Data-driven inverter-based Volt/VAr control for partially observable distribution networks

For active distribution networks (ADNs) integrated with massive inverter-based energy resources, it is impractical to maintain the accurate model and deploy measurements at all nodes due large-scale of ADNs. Thus, current models ADNs are usually involving significant errors or even unknown. Moreover, partially observable since only a few available pilot users. To provide practical Volt/Var control (VVC) strategy for such networks, data-driven VVC method proposed in this paper. Firstly, system response policy, approximating relationship between variables states monitoring nodes, estimated by recursive regression closed-form solution. Then, based on real-time newly updated convergence guarantee realized. Since solution embedded stage, closed-loop framework established. The effectiveness validated an unbalanced considering nonlinear loads where not rapid self-adaptive voltage regulation realized but also system-wide optimization achieved.