基于神经网络和Markov链的交通流实时滚动预测

将神经网络与Markov链理论应用于随机波动的交通流预测,提出一种交通流实时滚动预测方法TDFNM。该方法采用BP网络构建交通流基准预测曲线,使用SOM网络划分残差的Markov链状态,计算各状态加权中心及状态转移概率矩阵,以此预测未来状态,并以加权中点修正计算得到精度较高的预测值,同时实现实时滚动预测。采用方法TDFNM对实测交通流量进行仿真实验,结果表明,该方法比常规BP网络具有更高的准确性,而且具有较强的适应性。

Real-time Rolling Traffic Flow Forecasting Based on Neural Networks and Markov Chains

Neural networks and Markov chains were studied to forecast traffic flow, which has great randomness and fluctuation. A method (TDFNM) for real-time rolling traffic flow forecasting was proposed. This method used a BP neural network to establish a forecasting baseline. A SOM neural network was applied to divide the residual errors into different status for Markov chain. Then status weighting centers and status transition probability matrices were calculated. Subsequently, the status transformation was analyzed to determine the most possible status of the predicted value, and then the corresponding weighting center was used to revise the predicted value to achieve the more accurate one. Meanwhile, the traffic flow real-time rolling forecasting was realized. Using method TDFNM to conduct a simulation experiment on real data, the results demonstrate that this method is superior to the common BP neural networks in precision and has good adaptability to the dynamic traffic flow environment.