您的位置:首页  > 论文页面

BP神经网络与曲线回归耦合算法在交通量预测中的应用

发表时间:2010-03-31  浏览量:1430  下载量:512
全部作者: 范加冬,张令刚
作者单位: 中国矿业大学建筑工程学院
摘 要: 针对传统交通量预测方法中的局限性,采用BP神经网络与曲线回归耦合算法,用BP神经网络对历史数据训练,利用曲线回归对各因素待测年份之值进行预测,将各因素的预测值带入已训练好的BP神经网络中,即可得到未来交通量的预测值。通过对徐州解放路交通量实例预测,并和仅用BP神经网络、仅用曲线回归预测效果进行误差对比分析,结果显示此算法有较高的准确率。
关 键 词: 交通量预测;曲线回归;BP神经网络;曲线回归耦合算法;交通量
Title: Application of BP neural network and curvilinear regression coupling algorithm in traffic forecast
Author: FAN Jiadong, ZHANG Linggang
Organization: School of Architecture and Civil Engineering, China University of Mining and Technology
Abstract: For the limitations of the traditional algorithms that can not be reasonably used to predict the traffic, this paper adopted the BP neural network and curvilinear regression coupling algorithms to add the inadequate. First, BP neural network was used to train historical data and the curve regression was used to predict the various factors, and then the predicted values were substituted into the trained BP neural network which had been trained well, so the future traffic forecasts were got. By the example of Xuzhou Jiefang Road traffic forecast, and through error analysis, the results show that this algorithm has a higher accuracy.
Key words: traffic forecast; curvilinear regression; BP neural network; curvilinear regression coupling algorithm; traffic
发表期数: 2010年3月第6期
引用格式: 范加冬,张令刚. BP神经网络与曲线回归耦合算法在交通量预测中的应用[J]. 中国科技论文在线精品论文,2010,3(6):619-622.
 
0 评论数 0
暂无评论
友情链接