西安石油大学学报（自然科学版）2024，Vol.39Issue(1)：129-134,6.DOI:10.3969/j.issn.1673-064X.2024.01.016
基于CNN-LSTM混合神经网络的光伏发电量预测方法研究
Research on Photovoltaic Power Generation Prediction Method Based on CNN-LSTM Hybrid Neural Network
王登海 1安玥馨 2廖晨博 3马家园4
作者信息
- 1. 长庆油田分公司,陕西西安 710103
- 2. 长庆油田分公司新能源事业部,陕西西安 710103
- 3. 长庆油田分公司数字和智能化事业部,陕西西安 710103
- 4. 西安石油大学计算机学院,陕西西安 710065
- 折叠
摘要
Abstract
The photovoltaic power generation is influenced by factors such as weather conditions,the quality of photovoltaic inverters,and the cleanliness of photovoltaic modules,and among weather conditions the seasonal changes in have a significant impact on the pow-er generation.A photovoltaic power generation prediction method based on a hybrid model consisting of convolutional neural network(CNN)and long-short-term memory(LSTM)is proposed to solve the issue of inaccurate photovoltaic power generation prediction caused by temporal changes in weather in different regions.The spatial correlation between regions is established through CNN,and the temporal relationship between power generation data is captured through LSTM.The test results of photovoltaic power generation data from Hongmin Power Plant in Shenmu County and Greenergy Power Plant in Qingcheng County show that the proposed CNN-LSTM hy-brid neural network method for predicting photovoltaic power generation has high accuracy and stability,with a prediction accuracy im-provement of about 4.3％compared to the LSTM neural network method.关键词
光伏发电/模型预测/机器学习/CNN/LSTMKey words
photovoltaic power generation/model prediction/machine learing/CNN/LSTM分类
信息技术与安全科学引用本文复制引用
王登海,安玥馨,廖晨博,马家园..基于CNN-LSTM混合神经网络的光伏发电量预测方法研究[J].西安石油大学学报（自然科学版）,2024,39(1):129-134,6.
