Article Details

Abstract

Wind energy, as a renewable and clean energy source, is developing at a high speed. High-precision wind power generation forecasting can provide an important basis for power dispatching and mitigating grid-connection fluctuations. Based on the ARIMA model, this paper conducts ultra-short-term, short-term, and medium-to-long-term wind power generation forecasts to explore its feasibility and effectiveness. The wind power data of a wind farm from January 1, 2019 to December 31, 2020 is used for modeling and forecasting over different time lengths. On this basis, considering the periodic characteristics of the data and the influence of actual climate factors on power generation, the forecasting effect of the model is optimized. The experimental results show that considering the influence of periodic data and exogenous variables can significantly improve the forecasting accuracy of actual power. The RMSE of ultra-short-term, short-term, and medium-to-long-term forecasts are 7.16, 12.63 and 17.98, respectively.

Keywords

wind power generation; power generation forecasting; ARIMA; Pearson correlation coefficient

References

1. [1] Yang Chengsheng Short term wind power forecasting method for new wind farms [J]. China Science and Technology Information, 2022 (19): 104-106
2. [2] Chen Zihan, Teng Wei, Xu Xuefeng, etc Medium and Long term Wind Power Prediction Based on Graph Convolutional Networks and Wind Speed Differential Fitting [J]. China Electric Power, 2023, 56 (10): 96-105
3. [3] Chen Xiaowen, Su Pan, Wu Binrong, etc Research on Time Series Prediction Based on Improved Long Short Term Memory Networks [J]. Journal of Wuhan University of Technology: Information and Management Engineering Edition, 2022, 44 (3): 487-494+499
4. [4] Cao Junbo, Zhou Renjun, Deng Xuehua, etc Consider optimizing the ARIMA model differential frequency for wind power prediction [J]. Journal of Power Systems and Automation, 2019, 31 (1): 105-111
5. [5] Xiang Yong, Wang Yijun, Kong Yinke Large scale wind power prediction based on deep recurrent neural network [J]. Resource Information and Engineering, 2021, 36 (2): 136-139
6. [6] Li Xiangjun, Xu Gejian Wind power prediction method based on long short-term memory neural network [J]. Power Generation Technology, 2019, 40 (5): 426-433
7. [7] Han Peng, Zhang Xiaolin, Zhang Fei, etc Ultra short term wind power prediction based on AM-LSTM model [J]. Science, Technology and Engineering, 2020, 20 (21): 8594-8600
8. [8] Yu Zheng, Jin Bo, Jiao Yaoyi, etc Research on Short term Wind Power Prediction Based on Hybrid Deep Learning [J]. Journal of Wuhan University of Technology: Information and Management Engineering Edition, 2024, 46 (1): 170-174
9. [9] He Jianwei, Cao Yukun Long term wind power combination prediction method using LSTM-RF [J]. Journal of Shanghai Electric Power University, 2020, 36 (4): 341-350
10. [10] Ding Rui, Li Wei, Wang Ruozhou Based on SARIMA and LSTM combined prediction model [J]. Computer and Digital Engineering, 2020, 48 (2): 304-307+337
11. [11] Cheng Jie, Chen Ding, Li Chun, etc Ultra short term wind power prediction based on GWO-CNN BiLSTM [J]. Science, Technology and Engineering, 2023, 23 (35): 15091-15099

About the authors

Liu Jixiong (1971-), male, Han nationality, from Jingzhou, Hubei, engineer, undergraduate, research direction: air performance of fans; Zou Rui (1999-), female, Han ethnicity, born in Baoji, Shaanxi, holds a master's degree in data analysis and mining; Xu Siwei (1999-), male, Han ethnicity, from Tianmen, Hubei, holds a master's degree and specializes in data mining and machine learning.