基于双维度顺序填补框架与改进Kohonen天气聚类的光伏发电短期预测

为提高部分数据缺失情况下智能电网光伏发电功率预测精度,提出一种基于双维度顺序填补框架与改进Kohonen天气聚类的光伏发电功率预测模型。采用双维度顺序填补方法补齐缺失数据,基于完整数据分析光伏发电功率影响因素,建立改进Kohonen天气聚类模型,并利用S-Kohonen实现预测日天气类型识别,将聚类历史日数据与预测日气象数据作为输入,采用多种群果蝇优化广义回归神经网络(MFOA-GRNN)模型对预测日光伏发电功率进行预测。仿真结果表明,所提方法能有效提高预测精度,为实现数据缺失情况下智能电网光伏发电功率的精准预测提供了一种思路。

Short-term photovoltaic power forecasting based on double-dimensional sequential imputation framework and improved Kohonen clustering

To improve the accuracy of smart grid photovoltaic power forecasting in the absence of partial data, a forecasting method based on double-dimensional sequential imputation framework and improved Kohonen clustering is proposed. Firstly, using double-dimensional sequential imputation method to complement missing data, analyzing the influential factors of photovoltaic power based on the complete data, an improved Kohonen weather clustering model is established. Secondly, the weather type of the forecasting day is identified using S-Kohonen method. Finally, the MFOA-GRNN(Multi-swarm Fruit fly Optimization Algorithm and Generalized Regression Neural Network) model is used to forecast the daily photovoltaic power using the meteorological data of clustering historical day and forecasting day as input. The results show that the proposed method can effectively improve the forecasting accuracy and provide a way to accurately forecast the smart grid photovoltaic power under the condition of missing data.