Wind-Energy-Prediction-using-LSTM:使用LSTM进行风能预测的时间序列分析

  • b3_438826
    了解作者
  • 6.8MB
    文件大小
  • zip
    文件格式
  • 0
    收藏次数
  • VIP专享
    资源类型
  • 3
    下载次数
  • 2022-05-03 01:53
    上传日期
使用长期短期记忆(LSTM)进行风能预测 有关完整的详细信息,请阅读CSE 523项目报告.pdf。 介绍 由于风速/功率具有可再生性和环境友好性,因此在地球上受到越来越多的关注。 随着全球风电装机容量的Swift增加,风电行业正在发展为大型企业。 可靠的短期风速预测在风能转换系统中起着至关重要的作用,例如风轮机的动态控制和电力系统调度。 精确的预测需要克服由于天气条件波动而导致的可变能源生产问题。 风产生的功率高度依赖于风速。 尽管它是高度非线性的,但风速在特定时间段内遵循特定模式。 我们利用这种时间序列模式来获得有用的信息,并将其用于功率预测。 LSTM用于对数据执行不同的实验并得出结论。 结论 我们的目标是改善对使用风能发电的功率的预测,并且已经实现了将LSTM用作机器学习模型并对其进行模型优化。 我们还观察到,如果风速小于4 m / s,则系统生成的功率为零。 LSTM无法学习这
Wind-Energy-Prediction-using-LSTM-master.zip
  • Wind-Energy-Prediction-using-LSTM-master
  • Exp10-Prediction-Batch 419 with model optimization 1 week.ipynb
    152.8KB
  • Exp2-Estimation-8 look backs 60-40 split.ipynb
    166.2KB
  • Exp17-AL-Prediction Wind Approach 1 Batch 1 1 week ahead.ipynb
    211.5KB
  • Exp9-Prediction-Batch 386 with model optimization 48 hours.ipynb
    63.1KB
  • CSE 523 Project Report.pdf
    843.7KB
  • Exp18-AL-Prediction Wind Approach 1 Batch 1 1 month ahead.ipynb
    204.3KB
  • ~$ediction.docx
    162B
  • README.md
    1.8KB
  • Exp12-WA-Prediction Wind Approach 1 Batch 1 48 hours ahead.ipynb
    150.1KB
  • Exp14-WA-Prediction Wind Approach 1 Batch 1 month ahead.ipynb
    159KB
  • Exp3-Prediction-Batch 1-12 hour.ipynb
    43.1KB
  • Exp16-AL-Prediction Wind Approach 1 Batch 1 2 days ahead.ipynb
    208.1KB
  • ~WRL0005.tmp
    242KB
  • Exp15-AL-Prediction Wind Approach 1 Batch 1 24 hours ahead.ipynb
    206.7KB
  • WA_WIND_DATA.xlsx
    1.2MB
  • Prediction.pdf
    204.2KB
  • Exp6-Prediction-Batch 1-1 week.ipynb
    144.7KB
  • Exp11-WA-Prediction Wind Approach 1 Batch 1 24 hours ahead.ipynb
    176KB
  • Exp7-Prediction-Batch 1-1 month.ipynb
    207.6KB
  • AL_WIND_07_12.xlsx
    2.8MB
  • Exp4-Prediction-Batch 1-24 hour.ipynb
    45.7KB
  • Exp8-Prediction-Batch 264 with model optimization 12 hours.ipynb
    57.5KB
  • Exp5-Prediction-Batch 1-48 hour.ipynb
    55.3KB
  • Exp1-Estimation-8 look backs 70-30 split.ipynb
    146.1KB
  • Exp13-WA-Prediction Wind Approach 1 Batch 1 week ahead.ipynb
    169.8KB
内容介绍
# Wind Energy Prediction using Long Short Term Memory(LSTM) For complete details please read CSE 523 Project Report.pdf. # Introduction Wind speed/power has received increasing attention around the earth due to its renewable nature as well as environmental friendliness. With the global installed wind power capacity rapidly increasing, the wind industry is growing into a large-scale business. Reliable short-term wind speed forecasts play a practical and crucial role in wind energy conversion systems, such as the dynamic control of wind turbines and power system scheduling. A precise forecast needs to overcome problems of variable energy production caused by fluctuating weather conditions. Power generated by wind is highly dependent on the wind speed. Though it is highly non-linear, wind speed follows a certain pattern over a certain period of time. We exploit this time series pattern to gain useful information and use it for power prediction. LSTM is used to perform different experiments on the data and to form conclusion. # Conclusion We started with the aim of improving the predictions of power generated using wind energy and we have achieved that using LSTM as machine learning model and performing model optimization on it. ![alt text](https://github.com/ShashwatArghode/Wind-Energy-Prediction-using-LSTM/blob/master/Wind%20Prediction%20Result.JPG) We have also observed that if the wind speed is less than 4 m/s the power generated by the system is zero. LSTM is not able to learn this pattern as this is not the part which it can understand in time series analysis. So, if a hybrid new model is created which can work as the combination of Decision Tree/Random Forest and LSTM we can improve upon these results as well. # To read about all experiments please read CSE 523 Project Report.pdf
评论
    相关推荐