灰色模型和最小二乘支持向量机在短期负荷组合预测中的应用

Short-term Load Combination Forecasting by Grey Model and Least Squares Support Vector Machines

提出了结合数据预处理和灰色模型(GM)的最小二乘支持向量机(LS-SVM)短期负荷预测模型和算法.该模型在数据预处理的基础上,根据时刻T,通过缩小的样本集建立灰色模型,利用灰色模型的预测结果构建最小二乘支持向量机,最终,通过建立的最小二乘支持向量机对预测时刻进行预测.该算法不仅通过数据预处理策略提高了预测精度,而且避免了组合预测模型中权值选择问题.采用上述方法对河南电网负荷进行了预测分析,结果证明了该方法的有效性.

A short-term load forecasting method and the corresponding algorithm based on least squares support vector machines（ LS -SVM） ,grey model（GM） and data processing are proposed. Firstly, the model re-moves some outliers. Secondly, by reducing the sample set to build grey model based on time T, the model uses the predicted results of grey model to establish least squares support vector machines. Finally, the model predicts the load of prediction time by using the least squares support vector machines. The algorithm not only improves the prediction accuracy, but also avoids the combination forecasting model weights choice. The application of the proposed method to forecasting load of Henan power grid, shows that the proposed method is effective.