大比例分布式光伏接入电网的净功率预测仿真  

Net Power Prediction Simulation of Large Proportion of Distributed PV Access to Power Grid

作　　者：蔡超 韩俊 潘文婕 樊安洁 CAI Chao;HAN Jun;PAN Wen-jie;FAN An-jie(State Grid Jiangsu Electric Power Co.,Ltd.Economic and Technological Research Institute,Jiangsu Nanjing 210000,China)

机构地区：[1]国网江苏省电力有限公司经济技术研究院,江苏南京210000

出　　处：《计算机仿真》2025年第1期42-46,80,共6页Computer Simulation

基　　金：国网江苏经研院项目资助(SGJSJY00GHJS2400016)。

摘　　要：由于光照条件的变化和光伏组件的特性,大比例分布式光伏接入下光伏发电功率具有很强的波动性,使得净功率预测难度增加。因此,提出大比例分布式光伏接入下电网台区净功率预测方法。通过灰色关联分析法与余弦相似度结合方法,选取与待预测日功率数据最相似的历史日,将其作为时间预测模型的输入。使用AP聚类方法划分光伏电站群,获取待预测光伏电网台区所在的光伏电站群,并将其作为空间预测模型的输入。基于ATT-LSTM和LSSVM方法,设计具有时空关联性的光伏电网台区净功率组合预测模型,将获取的两项功率数据输入至组合模型内,实现分布式光伏接入下电网台区净功率预测。实验结果表明,所提方法的电网台区净功率预测准确度高、整体应用效果好。Due to the change of lighting conditions and the characteristics of photovoltaic modules,the photovoltaic power generation under large-scale distributed photovoltaic access has strong fluctuation,which makes it more difficult to predict the net power.Therefore,a method for predicting the net power of power grid area under large-scale distributed photovoltaic access is proposed.Through the combination of grey relational analysis and cosine similarity,the historical day most similar to the daily power data to be predicted is selected as the input of the time prediction model.An AP clustering method is used to divide the photovoltaic power plant groups,and the photovoltaic power plant groups where the photovoltaic power grid area to be predicted is located are obtained,which are used as the input of the spatial prediction model.Based on ATT-LSTM and LSSVM methods,a combined forecasting model of photovoltaic power grid area net power with temporal and spatial correlation is designed,and the obtained two power data are input into the combined model to realize the net power forecasting of grid area under distributed photovoltaic access.The experimental results show that the proposed method has high accuracy and a good overall application effect.

关 键 词：净功率预测 分布式光伏电站 聚类方法 

分 类 号：TP391.9[自动化与计算机技术—计算机应用技术]