摘要
以沈海高速罗长段隧道群交通事故为预测对象,以相对交通事故频数为预测目标,针对隧道群交通系统非线性、动态的特点,研究基于BP神经网络的交通事故非线性综合预测方法.结果表明,非线性综合预测方法能综合指数平滑模型、非线性回归模型、灰色马尔科夫模型等单一预测模型的有效隐含信息,提高交通事故的预测精度,为优化交通事故预测技术指出新方向.通过该方法预测,随着交通量的增加,未来3年(2016-2018年)罗长高速隧道群交通事故频数将呈现先下降后急剧上升的变化过程.
In order to forecast the traffic accidents of tunnel group in Luoyuan-Changle section of Shenyang-Hainan highway. Considering the characteristics of nonlinear and dynamic of the traffic system in tunnel group,the method of nonlinear integrated forecasting based on BP neural network was studied,and the relative traffic accident frequency was regarded as forecast target. The results show that nonlinear integrated forecasting model can integrate effective implicit information from single forecast model such as exponent smooth model,nonlinear regression model and Grey Markov model. The accuracy of forecast is improved by this method,and a new way is pointed out for optimize the traffic accident prediction technology. By this method,with the increase of traffic volume,the traffic accident frequency in this tunnel group will first decreased and then increased sharply in the next 3 years（ 2016-2018）.
出处
《福州大学学报（自然科学版）》
CAS
北大核心
2017年第2期226-230,共5页
Journal of Fuzhou University(Natural Science Edition)
基金
国家自然科学基金资助项目(5130405)
福建省交通运输科技项目(201526)
福建省教育厅科技项目(JAT160090)
关键词
隧道群
相对交通事故频数
非线性预测
综合预测方法
tunnel group
relative traffic accident frequency
nonlinear forecasting
integrated forecasting method
