基于聚类技术的光伏并网影响量化研究

大型光伏系统并网对电网性能的影响分析是安装大型光伏系统前的一项重要工作。可以采用基于时间模拟的方法进行影响量化分析，其优点是能够提供有关光伏系统发电量波动的信息，缺点是需要进行大量的分析和模拟，尤其是在采用以1 h为采样间隔的长历史数据时，不适用于影响研究，因此提出了一种有效利用数据的新方法。该方法利用聚类技术将具有相似特征的光伏功率输出段进行分组，每组选择1个有代表性的输出段，用以分析和模拟。该代表性的输出段可提供关于组内其他输出段的预期性能信息。最后，通过大型光伏系统的选址定容，验证了该方法的有效性和实用性。

Quantitative Study on Impact of Grid-connected Photovoltaic System Based on Clustering Technology

The impact analysis of grid-connected large-scale photovoltaic system on grid performance is an important job before installing the large-scale photovoltaic system. Quantitative impact analysis based on time simulation can be used. Its advantage is that it can provide information about the fluctuation of photovoltaic power generation， and the disadvantage is that a lot of analysis and simulation are needed， especially when long-term history data with sampling interval of 1-hour is used. It is not suitable for the impact studies， so a new method for effectively utilizing data is proposed. This method uses clustering technology to divide photovoltaic power output segments with similar cha-racteristics into groups， and selects one representative of the output segments for each group to do the analysis and simulation， can providing expected performance information about other output segments in the group. Finally， the validity and practicability of this method is verified by site selection and capacity determination of the large-scale photovoltaic system.