灰色关联与麻雀优化算法预测光伏发电量研究

Research on prediction of photovoltaic power generation by grey correlation and sparrow optimization algorithm

为了提高建筑屋顶小型分布式光伏电站并网可靠性,依据8.16 kW屋顶光伏电站2016―2021年1744 d的实际运行数据,并利用公开的天气资料,开展发电量预测研究。研究首先基于聚类划分天气数据,采用灰色关联法分析太阳辐射、温度、相对湿度等气象因素与光伏发电量的相关性;接着采用麻雀搜索算法优化预测模型,并与常规及采用粒子群优化的模型进行了对比。研究结果表明麻雀优化算法可有效提高预测精度,有助于电网提前安排调度,减少电网冲击。

In order to improve the grid-connected reliability of the small distributed photovoltaic power station on the roof of the building,the research on the prediction of power generation was carried out based on the actual operation data of 1744 d from 2016 to 2021 of the 8.16 kW roof photovoltaic power station and the public weather data was used.Firstly,the weather data were divided based on clustering,and the grey correlation method was used to analyze the correlation between meteorological factors such as irradiance,temperature,relative humidity and photovoltaic power generation,then sparrow search algorithm was used to optimize the prediction model,and compared with conventional and particle swarm optimization models.The research results show that the sparrow optimization algorithm can effectively improve the prediction accuracy,help to schedule the power grid in advance and reduce the impact of the power grid.