Probabilistic Forecasting for Photovoltaic Power Based on Improved Bayesian Neural Network

被引：0
|
作者：
Zhao K. [1 ]
Pu T. [1 ]
Wang X. [1 ]
Li Y. [1 ]
机构：
[1] China Electric Power Research Institute, Haidian District, Beijing
来源：
Dianwang Jishu/Power System Technology | 2019年 / 43卷 / 12期
基金：
国家重点研发计划; 中国国家自然科学基金;
关键词：
Bayesian neural network; Distributed photovoltaic; Probabilistic forecasting; T-distributed stochastic neighbor embedding;
D O I：
10.13335/j.1000-3673.pst.2019.1461
中图分类号：
学科分类号：
摘要：
Photovoltaic power forecasting is of great significance to power grid dispatching and operation. Traditional deterministic forecasting methods can hardly response to the fluctuation of photovoltaic power, bringing challenges to secure and stable operation of power grid. In this paper, a probabilistic forecasting method for photovoltaic power based on improved Bayesian neural network is proposed. The weight of the neural network is expressed in the form of probability distribution to improve the ability of the neural network to deal with randomness of photovoltaic output, and feature dimension is reduced according to the correlation of input and output data to suppress overfitting. Furthermore, fully connected neural network and one-dimensional convolution neural network are introduced into the Bayesian neural network to extract information from different input data to improve prediction accuracy. Simulations are performed on a realistic photovoltaic system. Results shows the proposed method has higher prediction accuracy than traditional deterministic forecasting models when the photovoltaic output fluctuates. Compared with other probabilistic forecasting methods, the proposed method provides a narrower power interval while maintaining high accuracy. © 2019, Power System Technology Press. All right reserved.
引用
收藏
页码：4377 / 4386
页数：9
相关论文
共 28 条
  • [1] Shu Y., Zhang Z., Guo J., Et al., Study on key factors and solution of renewable energy accommodation, Proceedings of the CSEE, 37, 1, pp. 1-9, (2017)
  • [2] Zhang S., Sun Y., Wei Z., Et al., Day-ahead optimal scheduling for active distribution network considering controllable photovoltaic system via probability model, Power System Technology, 42, 1, pp. 247-253, (2018)
  • [3] Lu Z., Huang H., Shan B., Et al., Morphological evolution model and power forecasting prospect of future electric power systems with high proportion of renewable energy, Automation of Electric Power Systems, 41, 9, pp. 12-18, (2017)
  • [4] Sheng H., Xiao J., Cheng Y., Et al., Short-term solar power forecasting based on weighted Gaussian process regression, IEEE Transactions on Industrial Electronics, 65, 1, pp. 300-308, (2018)
  • [5] Dai Q., Duan S., Cai T., Et al., Short-term PV generation system forecasting model without irradiation based on weather type clustering, Proceedings of the CSEE, 31, 34, pp. 28-35, (2011)
  • [6] Tan J., Deng C., Yang W., Et al., Ultra-short-term photovoltaic power forecasting in microgrid based on adaboost clustering, Automation of Electric Power Systems, 41, 21, pp. 33-39, (2017)
  • [7] Zhang D., Han X., Deng C., Review on the research and practice of deep learning and reinforcement learning in smart grids, CSEE Journal of Power and Energy Systems, 4, 3, pp. 362-370, (2018)
  • [8] Li J., Huang Q., Wei S., Et al., Improved deep learning algorithm based on S-BGD and gradient pile strategy and its application in PV power forecasting, Power System Technology, 41, 10, pp. 3292-3300, (2017)
  • [9] Wang J., Ran R., Song Z., Probability forecast of shortterm photovoltaic power generation based on improved depth restricted Boltzmann machine algorithm, Electric Power Automation Equipment, 38, 5, pp. 43-49, (2018)
  • [10] Sarkar S., Ajjarapu V., MW resource assessment model for a hybrid energy conversion system with wind and solar resources, IEEE Transactions on Sustainable Energy, 2, 4, pp. 383-391, (2011)
123