基于贝叶斯网络的进场航班排序队列稳定性研究

Research on sequencing stability for arrival flights based on Bayesian network

为了提高终端空域航空器到达时间预测的准确性和排序队列的稳定性,本文从分析影响航空器预计到达时间ETA的不确定因素出发,建立了同一机型航空器过同一边界点到达时间的概率模型,提出了一种通过实际飞行数据计算到达时间的概率方法,利用贝叶斯网络计算各影响因素对ETA影响程度.实验结果表明,用概率方法预测到达时间具有简单易操作且准确性高的特点.利用影响程度作为权重进行修正并量化了ETA的不确定性,有助于改善进场航班排序队列的稳定性.

In order to improve the accuracy of prediction of aircraft arrival time and the stability of queue in the terminal airspace, a probabilistic model for the same type aircraft crossing the same entry point was established through analysis of the main factors influenced the estimated time of arrival, and a probabilistic prediction method was proposed using a large amount of actual flight data. The uncertainty of estimated time of arrival was calculated using Bayesian network. The results show that the probabilistic method for estimated time of arrival could quantize the uncertainty exactly and be help to improve the stability of arrival queue.