摘要
针对高速公路事件检测这一非线性分类问题,提出一种基于概率神经网络的事件检测方法。阐述了概率神经网络的结构与训练算法,分析了事件对交通流的影响规律,并合理地选取了概率神经网络的输入量,用高速公路管理部门提供的样本数据进行了仿真研究。仿真实验表明,基于概率神经网络的事件检测方法具有学习速度快、泛化能力好、检测准确率高等优点,具有良好的应用前景。
Aiming at the problem of nonlinear classification in freeway incident detection,an incident detection method based on Probabilistic Neural Network （PNN） is proposed.The structure and training algorithm of PNN are formulated.Then the influence of an incident on the traffic flow is analyzed,and the PNN input variables are selected reasonably.Simulation research is carried out with the sample date provided by the freeway administrative office.Simulation experiments show that PNN incident detection method has such advantages as fast learning speed,good generalization ability and high detection rate.It is found to be potentially applicable in practice.
出处
《计算机工程与应用》
CSCD
北大核心
2007年第15期227-229,共3页
Computer Engineering and Applications
基金
广东省自然科学基金(the Natural Science Foundation of Guangdong Province of China under Grant No.06300326)
关键词
交通工程
高速公路
事件检测
概率神经网络
分类
traffic engineering
freeway
incident detection
probabilistic neural network
classification
