To ensure the safe and stable operation of the power system by accurately identifying and locating spontaneous explosive faults of overhead power line insulators

an object detection algorithm is proposed.Based on ConvNeXt and attention mechanism

this algorithm can effectively detect insulator spontaneous explosive faults from visible light images captured by equipment such as unmanned aerial vehicles and inspection robots.Firstly

ConvNeXt

a novel convolutional neural network

is used as the backbone network with 1∶1∶1∶3 stage module ratio to enhance the network′s ability to extract abstract semantic features.Secondly

a cross-stage local connection structure is used to reduce network parameters and computational complexity

thereby enriching network gradient connections.Finally

a convolutional attention mechanism CBAM is introduced to enhance the network′s ability to perceive target areas within complex backgrounds.Experimental results show that the improved insulator spontaneous explosive fault detection model achieves an average precision of 97.4%

an improvement of 1.4% over the baseline YOLOv7

effectively enabling the detection of spontaneous explosive faults in insulators.