Open AccessDesign of photovoltaic hot spot detection system based on deep learning
Author(s) -
Yong Ren,
Yi Yu,
Jing Li,
Wenhua Zhang
Publication year - 2020
Publication title -
journal of physics. conference series
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.21
H-Index - 85
eISSN - 1742-6596
pISSN - 1742-6588
DOI - 10.1088/1742-6596/1693/1/012075
Subject(s) - computer science , hot spot (computer programming) , photovoltaic system , robustness (evolution) , convolutional neural network , feature extraction , pattern recognition (psychology) , artificial intelligence , detector , convolution (computer science) , feature (linguistics) , algorithm , artificial neural network , engineering , telecommunications , biochemistry , chemistry , linguistics , philosophy , electrical engineering , gene , operating system
At present, it is difficult to detect the photovoltaic (PV) hot spots and the recognition efficiency is low. In this paper, an improved Single Shot MultiBox Detector (SSD) algorithm was designed for PV hot spot detection. The algorithm used the MobileNet network to replace the VGG16 convolutional neural network structure in the original SSD. This network is a depthwise separable convolution structure. Using it for feature extraction can reduce the number of parameters in the structure and achieve the purpose of speeding up the network. The experimental results show that the improved algorithm can detect the hot spots of PV array with good confidence, low detection rate and good robustness. Compared with the You Only Look Once(YOLO) algorithm and the original SSD algorithm, the detection speed is significantly improved, which verifies the effectiveness of the algorithm.