交叉口瞬时交通流量预测的自适应卡尔曼滤波模型

an Improved Adaptive Kalman Filter Model for Short-term Traffic Flow Prediction at Intersection

提出了一个交叉口瞬时交通量预测的带时间窗自适应卡尔曼滤波模型(Adaptive Kalman Filter Model,AKFM);详细阐述了模型的理论基础、建立过程、推导步骤及状态转移矩阵;并通过C++实现了AKFM和传统卡尔曼滤波模型(Traditional Kalman Filter Model,TKFM);最后,通过实际交叉口的检测数据对该模型与传统卡尔曼滤波模型进行了比较分析及评价。实验结果表明,AKFM模型是稳定有效的,预测精度优于TKFM。

The paper presents an adaptive Kalman filter model （AKFM） with time-window for short-term traffic flow prediction at intersection.Then it studies the theoretical basis,the modeling process,the derivation steps and state transition matrix and implements the AKFM and Traditional Kalman Filter Model（TKFM） by C ＋ ＋.Finally,in experimental study,the paper shows that AKFM is stable,adaptable and effective.It can promote the prediction efficiency and lower the relative error of prediction in limited time and prediction accuracy is better than TKFM.