摘要
在二阶BP神经网络基础上加以改进,提出一种快速二阶BP神经网络,并将把该方法成功地用于公路交通量的预测中,通过与其它方法的比较分析,得出快速二阶BP神经网络预测方法加快了收敛速度,提高了结果的准确度,为科学地预测公路交通量提供了有力依据。
A new kind of fast second order BP neural networks is presented in the paper based on the traditional BP neural networks. And this method is successfully tested by application in the forecast of road traffic volume. Compared with other methods of forecast, the new method is proved to improve the precision of forecast and reduces the iterative times, which provides a robust tool for the scientific forecast of road traffic volume.
出处
《中南公路工程》
2006年第2期48-50,88,共4页
Central South Highway Engineering
基金
"985工程"GIS与遥感的地学应用科技创新平台项目(105203200400006)
国家自然科学基金(40471105
40471106)
中山大学青年教师科研启动基金资助项目
关键词
快速二阶BP神经网络
交通量
预测
fast second order BP neural networks
road traffic volume
forecast
