基于需求不确定性的机场拥挤风险预测模型与方法

为了获得机场交通需求的概率分布及其变化规律,量化机场交通需求预测的不确定性,从需求不确定性角度分析了航空器进离港时刻对机场交通需求预测的影响,基于多个时段交通需求相互转化的不确定性,建立了多时段机场进离港交通需求概率分布模型.在此基础上,将进离港交通需求与进离港容量曲线相匹配,建立了机场拥挤风险预测模型,给出了具体求解过程与方法.亚特兰大机场实际航班运行数据的验证结果表明,机场概率需求预测值比确定型需求预测值更接近实际进离港交通需求值;与确定型拥塞预测方法的准确度60.0%相比,本文模型将拥挤预测提高到80%;用旧金山机场实际航班数据验证了本文方法的有效性,准确性达到87.5%,为机场拥挤管理提供了依据. 

Risk Prediction Model and Methodology of Airport Congestion Based on Probabilistic Demand

In order to obtain the probabilistic distribution and variation of the airport traffic demand for a future time interval and quantify the uncertainty of airport demand, the influence of arrival-departure timing on traffic demand prediction was analyzed from the viewpoint of uncertainty in traffic demand. Based on the uncertainty of transformation among traffic demands of multiple intervals, a probabilistic distribution model of airport arrival and departure capacity demand for multiple intervals was established. On this basis, a risk prediction model of airport congestion was developed by matching the departure traffic demand with the arrival-departure capacity curve. In addition, specific steps and method for solving the model were presented. The proposed models were verified using the real flight data of the Atlanta (ATL) airport. The results show that the departure traffic demand values by the probabilistic demand prediction are much more closer to the real demand values than by the deterministic prediction method. The risk prediction model and method could increase the accuracy of airport congestion prediction to 80%, in comparison to the 60% accuracy of the deterministic prediction method. The validity of the proposed method was also verified using the real flight data of the San Francisco (SFO) airport with an accuracy up to 87.5%. Therefore, the proposed method can provide a theoretic foundation for airport congestion management.