基于熵权时序模型的超短期风电功率预测

针对风电具有较强的随机性和波动性,传统的单一预测方法难以准确描述其规律且预测精度较低的问题,提出风速熵和功率熵的概念,在时间序列法的基础上分别采用基于风速和基于功率的预测方法,并根据风速熵和功率熵的计算结果动态设置预测点的权值,建立风电功率的熵权时序模型。算例分析结果表明,所提方法能有效提取风速及功率历史数据中的有用信息,提高超短期风电功率预测精度,预测结果的准确率和合格率均优于神经网络法、时间序列法和基于风速法。

Super Short term Wind Power Forecasting Based on Entropy Weight Time series Model

Because of the strong randomness and volatility of the wind power, the traditional single forecasting method is difficult to describe its rule accurately and the prediction accuracy is low. Aiming at this problem, the concept of wind speed entropy and power entropy were proposed. On the basis of time series method, an entropy weight time series model was established by using prediction method based on wind speed and power, respectively. And then the forecasting weights were set dynamically according to wind speed entropy and power entropy. The example results show that the proposed method can effectively extract useful information from wind speed and power historical data, improve super short term wind power prediction accuracy; furthermore, the prediction accuracy and percent of pass are superior to neural network, time series method and the method based wind speed.