空中交通流量的短期预测

A Model for Short-term Prediction of Air Traffic Flow with Holt-Winters Model

用霍尔特-温特斯模型对机场的短期空中交通流量进行预测,首先介绍了该模型的具体预测步骤,并采用面向对象的软件技术,设计开发了预测模型软件。该预测模型软件除了可以根据用户设定的平滑系数进行预测之外,还可以自动计算出预测效果较优的"最优平滑系数"。本文利用自1998年来首都机场各个月份的交通流量实际数据,对2006年的12个月份数据进行预测,经过与实际数据的比较发现相对误差在3%左右。结果表明,基于"最优平滑系数"的霍尔特-温特斯模型能较好的预测机场的短期空中交通流量,是一种为空中交通战略流量管理提供准确数据的有效方法。

The paper carried on the forecast of the airport short-term air traffic flow and described the concrete prediction steps with Holt-Winters model. According to this model, we have designed a software by adopting object software technology. The software could calculate the optimum smooth coefficient and accomplish comparatively accurate prediction to the air traffic flow of Airport ZBAA of 2006. Prediction results show that the error rate of the actual value and the test value is within about 3%, and the model is feasible.