基于小波分析的交通参数组合预测方法  

Combined method for traffic parameter prediction based on wavelet analysis

作　　者：冯金巧[1] 杨兆升[2] 孙占全[1] 张立东[1] 刘威[1] 

机构地区：[1]山东省计算中心,济南250014 [2]吉林大学交通学院,长春130022

出　　处：《吉林大学学报（工学版）》2010年第5期1220-1224,共5页Journal of Jilin University:Engineering and Technology Edition

基　　金：山东省科技攻关项目(2008GG10001018)

摘　　要：为了更加准确地预测交通参数的变化趋势,结合小波理论处理时变信息的优势,设计了一种在小波分析的基础上利用BP神经网络进行预测的组合预测方法。该方法采用小波理论与神经网络结合的策略,具有普适性,且比传统的基于小波分析的组合预测过程简单,为大运算量的实时应用提供了可能。利用实际数据对该方法进行了验证,并与对比算法进行了效果对比。结果表明,本文提出的方法预测效果较好,具有一定的实用性。Considering the superiority of the wavelet theory in dealing with the time-varying informations, a combined method was proposed based on the wavelet analysis and the BP neural network to predict the trend of the traffic parameters more accurately. The new method is characterized by general adaptability because of the strategy combining the wavelet theory and the neural network. It is simpler than the traditional wavelet-based combined method, providing the possibility of application in the real-time large amount computation. The method was verified by the practical traffic data and compared with the traditional equivalence. The results show that the proposed method leads to better prediction, is applicable for the real condition.

关 键 词：交通运输工程 交通参数预测 组合预测 小波分析 BP神经网络 

分 类 号：U491[交通运输工程—交通运输规划与管理]