Substation Insulator Inspection Robot System Based on Deep Learning

Substation plays an important role in power system As the key electrical isolation component in substation, the state of insulator is directly related to the safe operation of power grid In order to improve the inspection efficiency and safety of substation insulators, this paper presents a substation insulator inspection robot system based on ResNet18 model in order to improve the inspection efficiency and safety of substation insulators The system consists of visual inspection module, navigation module and main control system The visual inspection module adopts Hi3516 visual identification terminal to accurately inspect insulators; The navigation module uses lidar and Rising Sun X3 to collect map data and SLAM algorithm to realize accurate positioning and path planning; The main control system takes STM32 core board as the center and is responsible for motor control network communication and other tasks In addition, a man-machine interface is designed to display the inspection status and inspection results of the robot in real time Experiments show that the system has high accuracy and robustness in practical application, which provides technical support for automatic inspection of insulators in substations.