A real-time foreign object detection system for power transmission equipment based on convolutional neural network

Abstract

The invention relates to a system for detecting foreign matters (including kites, plastic bags and the like) of real-time power transmission equipment based on a convolutional neural network. The method comprises the steps that a large number of cameras are used for periodically taking pictures of power transmission equipment, the pictures are transmitted into low-cost embedded equipment with Raspberry Pi as a chip, whether foreign matter exists in the power transmission equipment or not and the positions of the foreign matter are detected in real time through a set of deep learning model trained on a server, and the information is transmitted back to the server. A deployed recognition model utilizes a deep separable convolutional neural network to extract image features, so as to optimizethe Faster R-for foreign matter features. And real-time efficient detection is carried out in the CNN method. Whether foreign matters exist in the power transmission equipment can be detected in realtime, a large amount of manpower is saved to detect the condition of the power transmission equipment on site, and stable operation of the power transmission equipment is guaranteed.

Description

technical field [0001] The invention relates to the fields of power systems and computer vision, in particular to a system for real-time detection of foreign objects in power transmission equipment based on a convolutional neural network. Background technique [0002] If there are foreign objects in the power transmission equipment (including kites, plastic bags and other things that should not be in the power transmission equipment), it may affect the stability of the power transmission, and even cause serious safety problems. It is very important to check and clean up these foreign objects in time necessary. At present, the inspection and maintenance of power transmission equipment mainly rely on on-site surveys, and human eyes can identify whether there is any abnormality. In recent years, due to the development of drone technology, drones can be used to take pictures, and then these photos can be screened by personnel , and thus save some manpower, but it still cannot m...

AI Technical Summary

Problems solved by technology

[0005] Faster r-cnn is a very mature method in the field of target detection, but this method requires a large amount of memory consumption, and low-cost embedded devices cannot support it well. The speed of running in the CPU of the device is very slow, which cannot meet the actual application requirements

[0006] In addition to Faster R-CNN, there are also some faster lightweight target detection models, but the recognition accuracy will decrease, especially for small objects, and many foreign objects on power transmission equipment are relatively small.

Images