基于SVM-Markov模型的短期风速组合预测

The short-term wind speed combined prediction based on SVM-Markov

文章先建立SVM风速预测模型对风速进行初步预测,再根据风速的Markov转换概率求取预测风速的置信区间及其中值,将置信区间中值与初步预测值按照一定的权值进行组合,并对权值进行自适应调整得到组合预测值。仿真结果表明,风速组合预测值比初步预测值具有更高的精度。

The SVM prediction model is established to develop preliminary prediction of wind speeds.Then,the confidence intervals of the predicted wind speeds and their medians were obtained according to the Markov conversion probability of the wind speed.Finally,the medians of the confidence interval are combined with the preliminary predicted value to calculate finial predicted wind speed by adaptive adjusting weights.The simulation results show that the combined predicted values of wind speed behave higher accuracy than the preliminary predicted values.