基于神经网络的交通控制诱导协同模型

针对传统交通控制与诱导模型及算法的不足，提出了具有中心协调系统（CCOS）的交通控制与诱导协同模型。利用数据融合技术将历史数据的短时交通预测、交通事件检测结果以及实时交通流数据设计面向交通动态的信息融合，并采用神经网络技术构建基于神经网络的交通控制诱导协同模型，同时对模型的参数进行了确定。。通过典型的路网进行仿真实验和对比分析，实验验证了该模型具有可行性和有效性。

In view of the imperfection of the traditional traffic control and guidance model and algorithms, a coordinated traffic control and guidance model based on central coordination system （CCOS） is proposed. In this model, the short- term traffic prediction results of the past, the traffic incident detection results and the real-time traffic flow data are used to design dynamic traffic oriented information fusion. Moreover, using the neural network technology, a traffic control and guidance coordination model based on neural network system is constructed, whose parameters are decided by the experiments. Finally, a number of typical local road networks are selected for simulation and comparative experiments. The experiments show this model is feasible and effective.