| 基于VMD-LSTM与误差补偿的光伏发电超短期功率预测 |
| |
| 引用本文： | 王福忠,王帅峰,张丽. 基于VMD-LSTM与误差补偿的光伏发电超短期功率预测[J]. 太阳能学报, 2022, 43(8): 96-103. DOI: 10.19912/j.0254-0096.tynxb.2021-0043 |
| |
| 作者姓名： | 王福忠 王帅峰 张丽 |
| |
| 作者单位： | 河南理工大学电气工程与自动化学院,焦作 454000 |
| |
| 基金项目： | 国家自然科学基金（61403284）; 河南省科技攻关项目（202102210295; 212102210146）; 河南理工大学青年骨干项目（2019XQG-17） |
| |
| 摘 要： | 光伏序列具有的较高复杂性对光伏发电功率的预测精度产生了极大影响,对此提出一种基于VMD-LSTM与误差补偿的光伏发电超短期功率预测模型。该模型第1阶段采用VMD算法将原始功率序列分解为若干个不同的模态,并对其建立对应的LSTM网络模型进行预测,通过对各模态的预测结果求和得到初始预测功率;第2阶段采用LSTM网络对误差序列进行误差补偿预测,然后将初始预测功率和误差预测功率求和得到最终预测结果。仿真结果表明,该预测模型对天气具有较高的适应性,预测精度达到97%以上。
|
| 关 键 词： | 光伏发电 功率预测 深度学习 长短期记忆 变分模态分解 误差补偿 |
| 收稿时间： | 2021-01-11 |
ULTRA SHORT TERM POWER PREDICTION OF PHOTOVOLTAIC POWER GENERATION BASED ON VMD-LSTM AND ERROR COMPENSATION |
| |
| Wang Fuzhong,Wang Shuaifeng,Zhang Li. ULTRA SHORT TERM POWER PREDICTION OF PHOTOVOLTAIC POWER GENERATION BASED ON VMD-LSTM AND ERROR COMPENSATION[J]. Acta Energiae Solaris Sinica, 2022, 43(8): 96-103. DOI: 10.19912/j.0254-0096.tynxb.2021-0043 |
| |
| Authors: | Wang Fuzhong Wang Shuaifeng Zhang Li |
| |
| Affiliation: | School of Electrical Engineering and Automation, Henan Polytechnic University, Jiaozuo 454000, China |
| |
| Abstract: | The high complexity of photovoltaic sequences has a great impact on the prediction accuracy of photovoltaic power generation. Therefore, an ultra-short-term power prediction model of photovoltaic power generation based on VMD-LSTM and error compensation is proposed. In the first stage of the model,the VMD algorithm is used to decompose the original power sequence into several different modes,and the corresponding LSTM network model is established for prediction,and the initial predicted power is obtained by summing the prediction results of each mode;In the second stage, the LSTM network is used to perform error compensation prediction on the error sequence,and then the initial prediction power and the error prediction power are summed to get the final prediction result. The simulation results show that the prediction model has high adaptability to the weather, and the prediction accuracy is over 97%. |
| |
| Keywords: | PV power generation power forecasting deep learning long short-term memory variational mode decomposition error compensation |
|
| 点击此处可从《太阳能学报》浏览原始摘要信息 |
|
点击此处可从《太阳能学报》下载免费的PDF全文 |
