基于ARIMA-SVR组合方法的航班滑出时间预测

Prediction of flight taxi-out time based on ARIMA-SVR combination method

针对现有大型机场采用的航班滑出时间预测方法精度不高而导致的场面交通情况拥堵、运行效率低等问题,提出差分自回归移动平均(autoregressive integrated moving average,ARIMA)模型与支持向量回归(support vector regression,SVR)模型组合的离港航班滑出时间预测模型。在滑出时间预测问题中采用时间序列分析方法,首先用ARIMA方法对数据进行拟合,完成ARIMA预测;其次将ARIMA模型预测结果的残差作为构建SVR模型的输入,通过SVR模型预测残差以补偿序列中的非线性变化;最后将2个模型预测结果合并得出结论。经过实例仿真分析可以看出,组合预测模型精度优于单一ARIMA预测模型,可将滑出时间的预测精度提高至90%,能够有效预测航班滑出时间。

The existing methods for estimating the taxi-out time of large airports do not have high accuracy,which brings about problems such as traffic jams and low operating efficiency.Based on this problem,the autoregressive integrated moving average(ARIMA)model and support vector regression(SVR)combined departure flight time prediction model were proposed.In the sliding-out time prediction problem,the time series analysis method was used for analysis.Firstly,the ARIMA method was used to fit the data to determine the various parameters that need to be used to complete the ARIMA prediction.Secondly,the residual error of the prediction result of the ARIMA model was used as the input to construct the SVR model,and the residual error was predicted by the SVR model to compensate for the nonlinear changes in the sequence.Finally,the prediction results of the two models were combined to reach a conclusion.According to the prediction of the residual value by SVR,it was combined with the prediction result of the ARIMA model as the final result.Through the simulation analysis of the example,it can be seen that the accuracy of the combined forecasting model is better than that of the single ARIMA forecasting model,and the forecasting accuracy of the slip-out time is increased to 90%,which can effectively predict the flight-out time.