基于改进EfficientNetV2算法的三相串联故障电弧检测

Three-phase series arc fault detection based on improved EfficientNetV2 algorithm

串联电弧故障在电气火灾中的成因中占据着重要的比例,在具有变频器的三相电机线路中,由于多种电弧电流复杂行为以及多种负载的存在,很难准确识别变频器后部线路中发生的串联电弧故障。为了解决这一问题,提出了一种改进的EfficientNetV2算法。搭建低电压三相电弧故障数据采集平台,采集了所需要的正常状态和故障状态的电流信号。为了充分利用机器视觉的优势,采用马尔可夫变迁场(MTF)将采集到的时域电流信号编码为图像。将MTF图像送入模型中进行训练和测试,该模型具有轻量级高效的通道注意力和双池化空间注意力,更加专注于电弧特征,提高网络性能。实验结果表明,该方法的准确率可达98.99%。

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%.