短时交通流智能混合预测技术

为了克服现有单项预测技术对不同交通流状况的局限性,提出一种新的短时交通流智能混合预测模型.该智能混合预测模型包括3个子模型:历史平均模型、人工神经网络模型和模糊综合模型.历史平均模型以历史数据为基础,利用一次指数平滑法良好的静态稳定特性,对交通流量进行预测.人工神经网络模型采用常见的由S函数神经元组成的1.5层前馈神经网络,由于人工神经网络具有强大的动态非线性映射能力,该模型对动态交通流量的预测具有较高的精度和满意度.根据上述2个单项模型的特点,为了充分利用它们对不同交通状况的适应性,进一步提高整体预测效果,采用模糊逻辑来综合这2个单项模型的输出,并把模糊综合模型的输出作为整个智能混合模型的最终交通流量预测值.实际应用结果表明,该混合模型的预测精度高于单项预测模型各自单独使用时的精度,发挥了2种模型各自的优势,是短时交通流预测的一种有效方法.

Intelligent hybrid forecasting technique for short-term traffic flow

In order to transcend the limitation of exisiting single forecasting technique on different traffic condition, a novel intelligent hybrid (IH) forecasting model for short term traffic flow was presented. The IH model included three sub models that were history mean (HM) model, artificial neural network (ANN) model and fuzzy combination (FC) model. Based on the historical traffic data the HM model forecasted the traffic flow by the single exponential smoothing method which held the good static stabilization character. Otherwise, the ANN model was a 1.5 layer feed forward neural network built by some common S fuction neurons. By means of the strong dynamic nonlinear mapping ability of ANN, this sub model could estimate the dynamic traffic flow in a very precise and satisfactory sense. In order to take advantage of the useful information of the HM model and the ANN model to improve the forecasting effect further, the two individual models reflecting practical problems from different respects were combined by fuzzy logic. The FC model mixed the two individual forecast results and its output was regarded as the final forecasting of the traffic flow. Practical application results show that the IH model, which takes advantage of the unique strength of the HM model and the ANN model, can produce more precise forecasting than that of two individual models. Thus, the IH model can be an efficient forecasting method to the short term traffic flow.