【摘要】 为了区分正常电流和串联故障电弧电流，提出了一种基于时域波形分析方法，根据时域电流中的过零点平肩部、尖峰部和随机性来识别故障电弧电流。通过搭建故障电弧试验平台的方式对不同负载(热水壶、电风扇、吹风机)进行故障电弧试验，采集正常和故障状态的电流数据，将电流数据按周期划分。由于故障电弧电流具有随机性、高频分量和零休时间，观察两种状态的电流波形的异同，经过多次试验提出将电弧电流的小波分解后的周期高频分量能量、周期零休区时间和周期尖峰次数作为故障特征的检测方法。更多还原

【Abstract】 In order to distinguish between normal current and series arc fault current, a time-domain waveform analysis method was proposed to identify arc-fault current according to the zero-crossing flat shoulder, peak and randomness in the time-domain current. By building a fault arc test platform, the fault arc test was carried out on different loads(kettles, electric fans, hair dryers), the current data of normal and fault states was collected, and the current data was divided by cycles. Due to the randomness, high frequency components and zero rest time of the fault arc current, the similarities and differences of the current waveforms of the two states were observed. After many experiments, a detection method was proposed that took the periodic high-frequency component energy, periodic zero rest time and periodic peak times after wavelet decomposition of arc current as fault features.更多还原