中文摘要：

光伏阵列中故障电弧类型分为串联型和并联型,针对不同类型的故障电弧需要采用不同的电路保护策略,研究光伏系统故障电弧特性、检测方法及故障电弧类型辨识方案对保证光伏系统运行的可靠性具有重要意义。为了能准确检测电弧信号,首先采集光伏系统直流母线电流信号,随后通过标准差提取信号时域特征、小波包分析提取信号频域特征,构建特征平面,根据信号在特征平面内的位置进行故障电弧检测;提出后断前检法（断开光伏系统直流母线,在断点前采集光伏阵列电压信号进行故障电弧再检测）进行串并联故障电弧区分。理论分析和实验结果验证了所提方法的可行性和正确性。

英文摘要：

The arc faults in Photovoltaic array are divided into series and parallel type, different circuit protection strategies are needed for different types of arc faults. The research on arc characteristics, detection methods and the arc type identification schemes of photovoltaic system is very important to ensure the reliability of photovoltaic system. In order to detect the arc faults accurately, the DC bus ctwrent signal of the PV system was analyzed. In this paper, by using the standard deviation to extract the signal time domain characteristics, the wavelet packet analysis was used to extract the frequency domain feature, and the time-frequency feature plane was constructed. The arc detection was carried out according to the position of the signal in the characteristic plane. By forcing the PV （photovoltaic） array to Voc and recheck arc noise to identify serial and parallel arc fault. The theoretical analysis and experimental results verified the feasibility and correctness of the proposed method.