基于最优邻域的动态加权混沌风速预测模型

WIND SPEED CHAOTIC PREDICTION MODEL BASED ON OPTIMAL NEIGHBORHOOD

提出了短期风速的混沌预测方法。首先利用关联积分法确定滞时和嵌入维数,重构风速时间序列的相空间。在此基础上,采用基于最优邻域的动态加权混沌预测模型进行风速预测。该模型综合考虑了邻近点权重和广义自由度,能够给出确定最优邻域的判定指标。实际计算中对2个测风点的数据进行了预测分析,结果表明,在合适的模型参数条件下,该方法可取得较好的预测效果,邻近点权重的引入确实提高了模型的预测精度。

Wind speed prediction is very important to wind power plants and power systems. Chaos theory and methods were used to discuss the wind speed prediction problem in the paper. Firstly, the time delay and the embedding dimension were calculated by correlation integral approach for reconstructing phase space of wind speed time series. Then wind speed chaotic prediction model of optimal neighborhood was proposed which gives overall consideration to the nearest neighbors＇ weights and generalized degrees of freedom, also an improved criterion for selecting optimal neighborhood. The practical calculation shows that the proposed model has superior predictive capability under the appropriate model parameters. After the nearest neighbors＇ weights were introduced, the model prediction precision was dramatically improved.