最小最大概率回归机在短时交通流预测中的应用

发布时间：2018-05-28 07:17

本文选题：交通工程 + 预测模型　；参考：《公路交通科技》2014年02期

【摘要】：为了提高短时交通流预测精度,更加精确地进行交通流规划和管理,引入一种新颖的基于最小最大概率回归机(MPMR)的短时交通流预测模型。针对北京某公路监测站实测的交通流数据集以及英国某地区实测的交通流数据集,利用基于MPMR的预测模型进行短时交通流预测,并与常规的基于神经网络、基于支持向量机(SVM)以及基于自适应神经模糊推理系统(ANFIS)预测模型的预测性能进行比较。试验结果表明,基于MPMR的短时交通流预测模型可以很好地跟踪实际流量值,在同等条件下相比常规预测模型的预测精度更优,验证了所提出模型的有效性。
[Abstract]:In order to improve the accuracy of short-term traffic flow prediction and to carry out traffic flow planning and management more accurately, a novel short-term traffic flow prediction model based on minimum maximum probability regression machine (MPMR) was introduced. In view of the traffic flow data set measured by a highway monitoring station in Beijing and the traffic flow data set measured in a certain area of England, the short-term traffic flow prediction is carried out by using the prediction model based on MPMR, and compared with the conventional one based on neural network. The prediction performance based on support vector machine (SVM) and adaptive neural fuzzy inference system (ANFIS) is compared. The experimental results show that the short time traffic flow forecasting model based on MPMR can track the actual flow value very well, and the prediction accuracy of the model is better than that of the conventional model under the same conditions. The validity of the proposed model is verified.
【作者单位】：兰州交通大学自动化与电气工程学院;
【基金】：甘肃省财政厅基本业务费项目(620026) 甘肃省硕导项目(1104-09)
【分类号】：U491.14

【参考文献】