基于压缩感知技术的大型光伏电站汇集系统故障定位研究

发布时间：2018-10-24 11:21

【摘要】：大型光伏电站汇集系统内线路分支密集,架空线与电缆共存,故障跳闸后巡线困难。而分布式与集中式并网光伏发电系统在拓扑和控制上有显著区别,造成相同电压等级的配电网故障定位方法在光伏电站内难以适用。为此,该文提出了一种符合站内实际运行控制的汇集系统故障定位方法。首先,结合现场逆变器控制,对站内故障电流特征进行分析。然后,利用光伏发电单元在不对称电压条件下不输出负序电流的特点,将故障后稀疏测点对应的节点负序电压方程和压缩感知理论相结合,利用贝叶斯压缩感知重构算法求解节点负序注入电流向量来进行故障定位。所提算法对测量数据无同步要求,所需数据较短。基于68节点系统的仿真数据表明,所提方法能够有效定位故障节点,且不受故障类型、过渡电阻等因素影响,抗噪能力强。
[Abstract]:In the collection system of large-scale photovoltaic power station, the branch of line is dense, the overhead line and cable coexist, the fault tripping is difficult to inspect the line. The topology and control of distributed and centralized grid-connected photovoltaic power generation systems are significantly different, which makes it difficult to apply the same voltage level distribution network fault location method in photovoltaic power plants. Therefore, this paper presents a fault location method of the gathering system in accordance with the actual operation control in the station. Firstly, the characteristics of fault current in the station are analyzed in combination with the field inverter control. Then, the negative sequence voltage equation of the node corresponding to the sparse measurement point after the fault is combined with the compression sensing theory, using the characteristic that the photovoltaic unit does not output the negative sequence current under the condition of asymmetric voltage. Bayesian compression perception reconstruction algorithm is used to solve the node negative sequence injection current vector for fault location. The proposed algorithm has no synchronization requirement for measurement data, and the required data is short. The simulation data based on 68 bus system show that the proposed method can locate the fault node effectively, and is not affected by the fault type, transition resistance and so on, and has strong anti-noise ability.
【作者单位】：新能源电力系统国家重点实验室(华北电力大学);国网苏州供电公司;
【基金】：国家重点基础研究发展计划项目(973计划)(2012CB215206) 国家自然科学基金(51407067) 中央高校基本科研业务费专项资金(2016ZZD01)~~
【分类号】：TM615

【参考文献】