有限信息下基于深度学习模型的小型分布式光伏电站功率预测

目的是解决小型分布式光伏电站在无气象站配备、无法测量气象变量(即太阳辐照度、温度、相对湿度等)的情况下,通过区域内光伏电站历史出力数据预测光伏发电的问题。基于有限信息,提出了两层的LSTM深度学习模型,对小型分布式光伏电站功率进行了预测,并对其超参数对其预测效果的影响进行了分析。此外,利用澳大利亚爱丽丝泉地区的分布式光伏电站数据来验证该模型的准确性,并与使用气象数据进行预测模型的效果进行了对比。结果表明,借助区域内光伏电站历史功率数据进行预测的效果良好,适用于无气象站情景下的光伏功率预测。

Power Prediction of Small Distributed Photovoltaic Power Plant Based on Deep Learning Model with Limited Information

The purpose of this paper is to solve the problem of predicting photovoltaic power generation in distributed photovoltaic power plant without meteorological information(ie,solar irradiance,temperature,relative humidity,etc.)To this end,this paper proposes a two-layer LSTM deep learning model to predict the power of distributed photovoltaic power plant based on limited information,and analyzes the impact of its hyper parameter on its prediction effect.In addition,this paper uses data from the distributed photovoltaic power plant in Alice Springs,Australia,to verify the accuracy of the model,and compares the effect with the prediction model using meteorological data.The results show that the accuracy of LSTM prediction based on the historical power data of photovoltaic power stations in the region is high,which is suitable for photovoltaic power prediction in scenarios without meteorological information.