Supervisor:
Southwest Ordnance Industry Bureau
Organizer:
Chongqing Ordnance Industry Society
Chongqing University of Technology
Chongqing University of Technology
Three phase series arc fault detection based on improved EfficientNetV2 algorithm
| DOI: | 10.11809/bqzbgcxb2024.03.039 |
| Keywords: | EfficientNetv2; attention mechanism; Markov transition field; three phase arc fault |
| Abstract: | The series arc fault occupies an important proportion in the causes of electrical fire. In the three phase motor line with frequency converter, it is difficult to accurately identify the series which arc fault in the back line of the frequency converter due to the complex behavior of various arc currents and the existence of various loads. In order to solve this problem, this paper proposes an improved EfficientNetV2 algorithm. To make full use of the advantages of machine vision, Markov transition field (MTF) is used to encode the collected time domain current signals into images. The MTF image is sent into the model for training and testing. The model has lightweight and efficient channel attention and dual pool spatial attention, which focuses more on arc features and improves network performance. The experimental results show that the accuracy of this method can reach 98.99%. |
| Published: | 2024-03-28 |