AlexNet+RCNN-based Model ImageRecognition and Knowledge Graphs Distribution Network Fault Alarm Method

Aiming at the accuracy and efficiency of fault alarms in distribution networks, we study the fault alarm method based on AlexNet+region convolutional neural network (RCNN) model for image recognition and knowledge graph. The classification and localization of faulty images are achieved by introducing the AlexNet model, which effectively performs image recognition and classification using deep learning techniques, and the RCNN model, which helps to locate and identify faulty regions. Combined with the construction and representation of the knowledge graph, the knowledge of fault types, causes and solutions is associated with the image recognition results to realize the intelligence of fault alarms. To evaluate the performance of the method, a distribution network fault dataset is constructed and the model is trained and tested in experiments. The experimental results show that the fault alarm method based on AlexNet+RCNN model and knowledge graph has achieved significant improvement in accuracy and efficiency. This study provides a useful reference for technical research and practical application in the field of fault alarming in distribution networks.