基于相似日重组的分布式光伏电源发电功率预测

Power Generation of Distributed Photovoltaic Power Generation based on Similar Days of Restructuring

在分布式光伏电源大规模接入的情况下,发电功率的预测问题受到越来越多的重视。本文在全面统计分析影响光伏电源发电功率的因素的基础上,提出可针对气象部门提供的预测日分时气象数据分别寻找相似数据点,并将历史数据重新组合成为新的相似日,再进行预测。本文以某区域分布式电源智能管控平台的实测数据为样本,运用BP算法进行训练和预测,取得了较好效果,尤其在天气突变时优势更为明显。

In the case of large-scale distributed PV power supply, the problem of power generation is becoming more and more important. On the basis of analyzing the factors that influence the power generation of photovoltaic power generation, the paper puts forward that the meteorological data can be used to predict the daily meteorological data. In this paper, we use the BP algorithm to train and predict the measured data of a regional distributed power supply intelligent control platform.