基于LSTM神经网络的中长期光伏电站发电量预测方法研究

Research on forecasting method of mid-and long-term photovoltaic power generation based on LSTM neural Network

准确估算光伏电站的中长期发电量对电网规划改进、调度优化、管理发展具有重要意义。然而,由于中长期发电量预测与短期出力预测存在显著差异,短期出力预测技术无法直接应用于中长期电量预测。文章提出一种基于模糊C均值聚类-随机森林算法FCM-RF和LSTM神经网络的中长期辐照度预测模型,进而提出间接预测分布式光伏电站发电量的方法。针对传统随机森林在数据差异性处理能力不足的问题,引入模糊C均值聚类算法对传统随机森林算法模型进行了改进。设计了LSTM神经网络,解决了"长时间周期依赖"问题。最后经实验验证,该分布式光伏电站中长期发电量预测模型每月预测平均误差百分数MAPE在3.5%上下波动,各电站年预测值在1.1%上下波动,预测效果较好。

Accurate estimation of medium and long term pv power generation is of great significance to grid planning improvement, scheduling optimization and management development.However, because of the significant difference between medium and long term power generation forecast and short term power generation forecast, short term power generation forecast technology cannot be directly applied to medium and long term power generation forecast. So in this paper, a medium-and long-term irradiance prediction model based on fuzzy C-means clustering-random forest algorithm FCM-RF and LSTM neural network is proposed, and then a method for indirect generation prediction of distributed photovoltaic power stations is proposed. In order to solve the problem that the traditional random forest can not deal with the difference of data, the fuzzy C-means clustering algorithm is introduced to improve the traditional random forest algorithm model.In order to solve the problem of "long-term cycle dependence", build an LSTM neural network model. Finally, the experiment verifies that the monthly average error percentage MAPE of the distributed pv power generation prediction model fluctuates around 3.5%, and the annual predicted value of each power station fluctuates around 1.1%, showing a good prediction effect.