基于自适应延时神经网络的短时交通流预测

为了提高交通流诱导系统的性能,研究了一种基于自适应延时神经网络算法的短时交通流量预测模型,该算法与传统的神经网络方法相比在神经网络的结构和神经网络的训练方法两个方面进行了改进,更适用于预测具有不确定性、非线性、动态性等特征的短时交通流,同时用Matlab7.0编写程序对算法进行了仿真实现,根据仿真结果的分析验证了算法时实性和精确性.

Forecast of Short Time Traffic Flow Based on Adaptive Time-Delay Neural Network

For improving the performance of Traffic How Guidance System,research of short time traffic flow forecast based on The Adaptive Time-Delay Neural Network algorithm,compared to traditional Neural Network algorithm,it has been improved in configuration and training method of Neural Network,ATNN is the same with uncertain characteristics,nonlinear and dynamic of the traffic flow.Simulation of the algorithm is carried on using Matlab7.0 procedure and the analysis of the result confirms the effectiveness and accuracy of the algorithm.