基于组合模型的短时交通流预测  

Short-Term Traffic Flow's Forecasting by Fusing Wavelet Neural Network and Historical Trend Model

作　　者：蒋亚平[1] 郭俊亮[1] 赵军伟[1] 

机构地区：[1]郑州轻工业学院计算机与通信工程学院,郑州450002

出　　处：《现代计算机》2013年第5期26-29,共4页Modern Computer

摘　　要：在交通管理中,为了解决城市交通道路拥塞问题,提出一种可靠稳定的短时交通流预测算法。基于小波神经网络和历史趋势两种方法建立单性预测子模型,提出用小波神经网络作为最优非线性组合模型的逼近器,建立组合预测模型,实现对单项预测子模型结果的融合,通过MatLab平台对实测交通流量数据进行了仿真,结果表明,该种组合预测在短时交通流量预测方面是切实可行的。In order to take measures in advance to avoid traffic congestion or traffic jamming that possibly exists in traffic management,proposes a short-term traffic flow forecasting algorithm by fusing wavelet-neural network and historical trend model. Uses the wavelet-neural network and historical trend model to establish the single forecasting sub-model separately. Then uses waveletneural network net is used as the approximate of the optimal nonlinear combination model to establish the combination forecasting model that can fuse the predicted value from the single forecasting sub-model, abtains the final forecasting result is obtained. By using MatLab platform, realizes the simulation based on the actual traffic flow data, and the results show that the combination forecasting technique is practical and feasible.

关 键 词：智能交通 短时交通流 小波神经网络 历史局势模型 

分 类 号：U491.112[交通运输工程—交通运输规划与管理]