基于小波变换遗传过程神经元网络的交通流预测  

Traffic flow prediction based on wavelet transformation epigenetic process neural networks

作　　者：高为 

机构地区：[1]广东广珠西线高速公路有限公司,广东佛山528305

出　　处：《山西建筑》2014年第3期160-162,共3页Shanxi Architecture

摘　　要：针对短时交通流时间序列的缺点,应用小波变换理论,将含有综合信息的时间序列分离为低频确定信号和高频干扰信号,用遗传过程神经元网络分别进行预测,得到了原时间序列的实际预测结果,通过实测数据验证表明,该预测方法具有较好的预测精度。In light of defects of short-time traffic flow time series, the paper applies wavelet transformation theory, divides eomprehensive time series into low-frequency determination signal and high-frequency disturbing signal, carries out a prediction by using epigenetic process neural networks, and obtains the actual prediction results of original time series. As a result, the actual testing data proves that, the prediction method has better prediction accuracy.

关 键 词：短时交通流预测 小波变换 过程神经元网络 

分 类 号：U491[交通运输工程—交通运输规划与管理]