The Research and Practice on Self-healing Application of Low-voltage Distribution Network Based on Cloud-Edge-End Collaboration

To address the challenges of low automation levels and inefficient operation and maintenance in low-voltage distribution networks, this study proposes a self-healing strategy based on a cloud-edge-end collaborative architecture. Initially, fiber-optic communication technology is employed to establish an interconnected low-voltage grid infrastructure, integrating Dianhong intelligent IoT switches to achieve the “telemetry, telecommunication, telecontrol and teleadjustment” quadruple remote function system at the low-voltage level. Subsequently, an interactive low-voltage station area diagram is developed within the edge-side system, innovatively extending medium-voltage distribution network mapping functionalities to low-voltage applications. Through refining the tree structure automatic layout algorithm, we realize topology visualization and remote control capabilities. Furthermore, a self-healing control mechanism integrating “cloud-side global optimization, edge-side real-time decision-making, and end-side rapid execution” is established. This framework incorporates a dynamic strategy inference engine for fault feature matching and multi-objective optimization decision-making, ultimately formulating comprehensive low-voltage self-healing strategies. Field validation conducted in Jingxia Village, Huangpu District, Guangzhou demonstrates that the proposed approach significantly reduces fault processing time from 45 minutes to seconds-level while substantially enhancing power supply reliability. This research provides a replicable technical pathway for intelligent low-voltage distribution network development and contributes practical experience to the standardized advancement of new power systems.