并网光伏系统孤岛检测策略研究

发布时间：2018-07-25 11:43

【摘要】：光伏系统的孤岛运行会导致电能质量恶化、电力系统设备损坏等问题,国内外的相关标准均规定光伏系统需要配备孤岛保护装置。传统的孤岛保护研究主要集中在单逆变器系统,对于多逆变器并联的光伏电站的孤岛保护研究甚少,传统的孤岛保护策略应用于光伏电站时则会存在孤岛检测失效的问题,需要研究适用于光伏电站的孤岛保护策略。论文根据光伏电站的实际系统结构建立了简化的光伏电站等效仿真模型,以此作为仿真验证的基础。分析了影响并网点处电压幅值和频率在光伏电站孤岛运行前后变化的因素,得出了传统被动式孤岛检测法的失效机理。以频率-无功反馈法为例,论证了传统主动法应用于光伏电站孤岛检测时,由于反馈系数的选择、扰动信号互相干扰等原因而会检测失效。论文提出了适用于光伏电站的基于SVG调节无功功率的孤岛检测法,在光伏电站输出的有功功率低于阈值时,以光伏电站输出到并网点的功率因数在指定区间循环波动为约束量,调节SVG输出的无功功率主动形成光伏电站和本地负载的无功不匹配,SVG输出的无功功率不影响光伏电站的并网运行,在光伏电站孤岛运行时,无功功率不匹配会导致并网点频率迅速超出阈值,可以以此快速可靠地检测出孤岛运行。为了达到避免人工设定参考量阈值、缩小孤岛检测盲区的目的,论文提出了适用于光伏电站的一种基于Adaboost智能算法的孤岛检测法。Adaboost算法利用若干低精度分类器组合成高精度分类器,学习由光伏电站不同运行情况下的特征量数据组合成的学习数据集,生成一个判断系统当前处于并网运行或孤岛运行状态的二分类模型,利用与学习数据集同源的检验数据集校验该模型的分类精度,将系统当前运行状态下的特征数据输入该模型即可判断系统是否处于孤岛运行。该方法不需人工设定参考量阈值,用多参考量同时判别孤岛,具有很高的检测精度。
[Abstract]:The isolated island operation of the photovoltaic system will lead to the deterioration of the power quality and the damage of the power system equipment. The relevant standards both at home and abroad stipulate that the photovoltaic system needs to be equipped with isolated island protection. The traditional island protection research is mainly concentrated in the single inverter system, and the Research on the island protection of the photovoltaic power station with multiple inverters is very small. When the isolated island protection strategy is applied to the photovoltaic power station, there will be a problem of isolated island detection failure. It is necessary to study the island protection strategy suitable for the photovoltaic power station. The paper establishes a simplified equivalent simulation model of the photovoltaic power station based on the actual system structure of the photovoltaic power station, which is the basis of the imitation true verification. The factors of amplitude and frequency variation before and after the operation of the photovoltaic power station before and after the island operation are obtained. The failure mechanism of the traditional passive island detection method is obtained. Taking the frequency reactive power feedback method as an example, it is demonstrated that the traditional active method is applied to the island detection of the photovoltaic power station, and the failure is detected due to the selection of the feedback coefficient and the interference of the disturbance signal to each other. In this paper, an isolated island detection method for reactive power based on SVG for photovoltaic power station is proposed. When the active power output of a photovoltaic power station is lower than the threshold, the power factor output to the node in the photovoltaic power station is circulated as a constraint in the specified range, and the reactive power of the SVG output is actively adjusted to form the photovoltaic power station and the local load. The reactive power of the SVG output does not affect the grid operation of the photovoltaic power station. When the photovoltaic power station is running on the island, the reactive power mismatch will lead to the rapid and reliable detection of the island operation. A.Adaboost algorithm based on Adaboost intelligent algorithm for photovoltaic power station is proposed in this paper. A high precision classifier combined with a number of low precision classifiers is used to learn a learning data set composed of the characteristic data of different operating conditions of the photovoltaic power station, and a judgment system is currently in the parallel operation. The two classification model of the running state of the isolated island is used to check the classification accuracy of the model by using the test data set homologous with the learning data set. The model can be used to judge whether the system is in the isolated island. The method does not need to set the threshold value manually and discriminating the island with the multi reference amount simultaneously. It has high detection precision.
【学位授予单位】：华北电力大学(北京)
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：TM615

【参考文献】