光伏系统串联直流电弧故障识别方法与定位策略研究

Series DC arc fault is a great threat to the safety of photovoltaic systems (PV systems). Many have proposed methods for detecting such arcs based on signals within frequency band of 40kHz-100kHz. It can cause refused tripping or nuisance tripping by the affection of background noise signals. Moreover, it can’t implicate the fault location. Those problems are studied in this project, which has important theoretical significance and application prospects..The mechanism of phenomenon of arc fault is studied, and then the mathematical models of arc I-V and arc radiation are established separately. Based on the study of PV modules, an arc fault model in PV system level is built. The affection of operating conditions of PV system on characteristics of arc I-V and arc radiation are investigated. Then, the key features of arc fault are extracted, which can distinguish arc signals from background noise signals. Furthermore, a multi-information fusion algorithm based on electromagnetic radiation signals and voltage and current signals is studied, which can overcome the shortcomings of single information source of conventional method and improve discrimination accuracy. Finally, the mechanism of propagation of arc fault radiation is analyzed. And a fault location strategy is proposed based on multi-antenna array to fulfill the fault location accurately.

光伏系统串联直流电弧故障是威胁系统安全运行的主要因素。目前普遍采用40kHz-100kHz的高频电弧电流分量进行电弧故障识别，由于受背景干扰影响，其检测精度不高，且难以实现故障准确定位。本项目以光伏系统串联直流故障电弧为研究对象，解决其中存在的科学问题，具有重要的理论意义和应用前景。.研究电弧特征现象的发生机理，建立电弧电压电流与电磁辐射的数学模型，并结合光伏系统各组件数学模型的研究，建立系统级电弧故障模型，解决电弧故障特性研究中的系统建模问题；研究光伏系统运行工况对电弧电压电流与电磁辐射特征参量的影响，提取电弧故障区别于背景干扰的本质特征，建立直流电弧故障特征库，进而研究串联直流电弧故障的多特征参量信息融合方法，进行电弧故障识别，提高故障识别准确率；分析电弧高频电磁辐射的产生与传播机理，研究基于天线阵列的电弧故障定位策略，实现故障点准确定位，为光伏系统安全稳定运行奠定基础。

光伏系统串联直流电弧故障是威胁系统安全运行的主要因素。目前普遍采用40kHz-100kHz的高频电弧电流分量进行电弧故障识别，由于受背景干扰影响，其检测精度不高，且难以实现故障准确定位。本项目以光伏系统串联直流故障电弧为研究对象，解决其中存在的科学问题，具有重要的理论意义和应用前景。.搭建了光伏直流电弧故障试验平台，开展了阻性与光伏逆变器负载条件下试验，建立了电弧故障数据库。基于电弧稳态信号Cassie模型与电弧暂态信号高斯白噪声模型，建立了单相光伏直流电弧故障系统级仿真模型，解决电弧故障特性研究中的系统建模问题； .研究了阻性负载与光伏逆变器负载条件下故障电弧电流时频域特征，提取了电弧故障特征量，时域特征量包括电弧电流均值、标准差以及极差；频域特征量为故障电弧电流50kHz~100kHz频段高频分量。分析了光伏阵列输出特性、放电间隙等对故障电弧电导的影响。采用电弧电流高频分量功率谱密度的AR模型参数作为特征量，提出了一种基于AR模型的光伏直流故障电弧检测方法。.分析了电弧故障电压电流与电磁辐射信号时频域特征以及电弧电流变化率、测量距离等对电弧电磁辐射信号强度的影响。提取了电弧故障电磁辐射特征量，提出了一种故障电弧多特征量融合检测方法，提高故障识别准确率。.提出了基于宽频带分形天线与Vivaldi天线的电弧电磁辐射测量方法，其信号带宽可达210MHz~800MHz。构建了电弧故障电磁辐射测量天线阵列，提出了基于RSSI混合模型与增强数据神经网络的故障电弧定位方法，实现了故障点定位，为光伏系统安全稳定运行奠定基础。