No abstract

Wind power generation is affected by weather and historical wind power, which presents the characteristics of instability and high volatility. Most wind power prediction models ignore physics information. In this paper, a novel combined predicting model that simultaneously considers physics information and historical information is presented to address the drawbacks of existing models. First, the physical characteristics of wind speed, wind direction, and temperature are obtained by Deep Neural Network(DNN), and time-series characteristics from historical wind power are extracted by Long Short-Term Memory(LSTM). Then, the physical features and the time-series features are fully connected for feature fusion to obtain the final time-series physical features. Finally, the short-term wind power prediction is performed according to the obtained merged features. Experimental results demonstrate that the DNN-LSTM model proposed in this paper achieves high accuracy and stability, and provides technical support for wind power system dispatch.

No abstract

Wind power generation prediction plays an important role in the safety and economic operation of the power system. There are many parameters recorded in wind farm data, such as wind power, wind speed, wind direction, and so on. Traditional wind power prediction modeling methods lack the mining of these parameter data and fail to make good use of some potential physical information. To address this challenge, this paper proposes a multiview neural network learning framework to predict wind power. One is the data attribute view of wind power, which uses the historical data feature of the wind power itself to learn the future wind power feature. The other is the physical attribute view of wind power, which uses the physical attribute features associated with the wind power definition to learn the future features. Then all the learned features are jointly fused to predict the future wind power values. In addition, an uncertain factor is proposed and computed, which is inspired by the wind power formula and usually associated with internal and external environment perturbation. All time series features are input into the gated recurrent unit neural network to form a hybrid neural network framework for wind power prediction. Experimental results under the measured condition and the standard condition of wind farms demonstrate the effectiveness of the proposed method.

No abstract