论文部分内容阅读
为了克服单一的航班延误预测模型在预报时的缺陷,在分析航班延误的特点的情况下,提出了一种基于危险模式和灰色预测组合的新型航班延误预测的方法。对这两种预测方法的结果采用加权组合预测的方式来预测航班延误的趋势变化,预测结果是单一预测模型的加权和,加权系数动态确定。最后通过国内的某枢纽机场的航班延误情况进行了验证。实验表明该模型可以不受某一较差的预测模型影响,从而有较好的预测效果。
In order to overcome the shortcomings of a single flight delay forecasting model in forecasting, this paper proposes a new flight delay forecasting method based on combination of dangerous mode and gray forecasting in the analysis of the characteristics of flight delays. The results of these two forecasting methods adopt the method of weighted combination forecasting to predict the trend change of flight delays. The forecasting result is the weighted sum of the single forecasting model and the weighting coefficient is dynamically determined. Finally, the flight delays at a hub airport in China were verified. Experiments show that the model can not be affected by a poor prediction model, which has a better prediction effect.