Data-model hybrid-driven supply restoration of active distribution networks with SOP

Increasing distributed generator (DG) penetration poses significant challenges in restoration performance of active distribution networks (ADNs). With dramatic development of flexible interconnection technology, power supply level of ADNs can be further improved by electronic devices such as soft open point (SOP). Traditional model-based restoration methods may mismatch with rapid regulation capability of SOP and be difficult to adapt to frequent DG fluctuations under fault situations. Thus, it is critical to fully explore the potential of SOP through data-driven method. To cope with these issues, a data-model hybrid-driven supply restoration method is proposed for ADNs with SOP. In day-ahead stage, model-based method is utilized to generate contingency set and guidance set. After fault occurrence, topology reconfiguration is determined by contingency set. Based on measurement feedback, data-driven load restoration is achieved with the rectification by guidance set. The case studies are conducted on a practical distribution network with four-terminal SOP. Results show that the proposed method can fully explore benefits of SOP to adapt to frequent DG fluctuations and promote restoration performance of ADNs.