Digital technologies for real-time 3D modeling and autonomous UAV inspection in complex distribution networks: A review

This paper systematically reviews the research status, key challenges and development trends of real-time 3D modeling and fully autonomous inspection technology for unmanned aerial vehicles (UAVs) in complex distribution networks. The article points out that traditional inspection methods have problems such as low efficiency, many blind spots, and low recognition accuracy. However, unmanned aerial vehicle technology, combined with LiDAR, oblique photography, and multi-source sensor fusion, can build centimeter-level real-scene 3D models, achieving automatic generation of flight routes and high-precision defect recognition. Meanwhile, the construction of the digital twin platform provides support for intelligent diagnosis and decision-making. The article focuses on analyzing technical difficulties such as environmental complexity, the accuracy and real-time performance of 3D modeling, autonomous navigation and obstacle avoidance, multi-source data fusion and communication support, and compares the advantages and disadvantages of various key technologies. Finally, the development directions of future technologies in multi-machine collaboration, edge intelligence, standardization and large-scale application were prospected, providing theoretical references and practical paths for the intelligent operation and maintenance of distribution networks.