短时交通流预测的多层递阶方法

短时交通流预测在城市交通控制和管理中起着十分重要的作用。然而,目前很多预测模型均假定模型的参数是不随时间变化的,这与实际不符,从而影响了预测的精度。本文提出采用多层建模与灰色建模的综合方法预测短时交通流。该方法把预测问题分为两部分:一是预测模型参数的预测;二是根据参数预测值的交通流预测。其中,对模型参数运用灰色理论预测方法。实例分析表明,本方法有较好的预测精度和实用价值。

Multi-Store Progressive Step Method of Short Time Traffic Flow Forecast

Short time traffic flow forecast plays very important role in urban traffic constrol and management. Howe ver, now many forecast models assume that the parameters of the models are not changed with time, so it not conforms to the practice, therefore, it influences on the forecast precision. The paper puts forward to forecast short time traffic flow by the comprehensive method of multi-store building model and gray building model. This method divides the forecast problem into two parts: one is forecast of model parameters, other is forecast of traffic flow in accordance with parameter forecast values. Among them the model parameters shall adopt gray theory forecast method. The example analysis shows that the method has the better forecast precision and useful value.