分解组合模型在短期燃气预测中的应用

Decomposition-combined model in the short-term prediction of gas

在对城市燃气负荷数据特性进行分析的基础上,提出了针对城市燃气负荷量短期预测的思想即分解-组合预测模型,同时提出了三种分解方法对分解-组合预测模型进行了验证。首先在建模之前运用数据挖掘的方法对原始数据集进行了离群点挖掘与修正;其次,为了验证准确性,将三种方法的预测结果与其他单一、组合模型预测结果进行对比;最后为了验证该模型的有效性、适用性,对特殊日期、天气和其另一组燃气负荷量数据集进行了建模和预测,通过对预测值和实际值的误差分析,实验结果进一步验证了分解-组合模型的适应性和准确性。

There are many forecasting methods in the short-term forecasts of city gas, however, a variety of methods are only suitable for a data set or only applicable in the special case of short-term forecast. In this paper, the authors put forward the idea of short-term forecast for the city gas load decomposition-combination forecasting model on the basis of the analysis of urban gas load data features. At the same time ,put forward three decomposition-combination forecasting model. First of all, before modeling, use data mining outliers mining and correction of the original data set. Secondly, in order to verify the accuracy of the prediction results of the three methods of text that will be submitted and other single combination predicted results were compared. In order to verify the validity and applicability of the model, modeling and forecasting on the special dates and special weather as well as an- other set of gas load data sets. Analysis on the predicted value and the actual value of the error, the experimental results further validate the proposed decomposition-combination of adaptability and accuracy of model.