摘要
对现有公路交通异常事件自动检测算法进行介绍,在此基础上提出从人工神经网络角度方向建立检测模型,摆脱以往使用仿真数据研究的局限,利用广深高速实测交通异常事件数据对BP神经网络(Back Propagation NN)在交通异常事件自动检测算法中的应用进行研究。结果表明BP神经网络算法具有检测率高、误报率低和检测速度快等优点。
Based on reviewing the existing algorithms used in free way automatic incidents detection（AID） algorithm,an artificial neural network detection model is proposed.Eliminating the limit of researches with simulation date and using the field traffic incidents date of GuangZhou-ShenZhen Highway ,the back propagation neural network algorithm applied in automatic incident detection is researched.The result shows its advantages,such as its higher detection rate ,lower false alarm rate and shorter mean detection time.
出处
《交通科技与经济》
2010年第6期47-50,共4页
Technology & Economy in Areas of Communications
基金
广东省科技计划资助项目(2009B010800052)
关键词
交通异常事件
自动检测算法
BP神经网络
incident detection
automatic detection algorithm
back propagation neural network
