论文部分内容阅读
为了精确预测路段出行时间,分析了国内外基于多数据源的路段出行时间预测方法的优缺点,应用自适应卡尔曼滤波算法,通过融合环形线圈检测器数据和浮动车数据,建立了路段出行时间估计模型,在交通高峰期和事故情况下,比较了采用基于环形线圈检测器、浮动车和自适应卡尔曼滤波3种出行时间预测方法预测路段出行时间的平均绝对百分比误差。比较结果表明:基于自适应卡尔曼滤波算法融合了来自环形线圈检测器和浮动车的数据,预测值更接近实测值,预测精度高。
In order to accurately predict the travel time of road segments, the advantages and disadvantages of travel time forecasting methods based on multiple data sources at home and abroad are analyzed. By using adaptive Kalman filter algorithm and by integrating the data of loop coil detectors and the floating car, the travel time of road segments The model estimated the average absolute percentage error of the travel time of the road segment by using three kinds of travel time forecasting methods based on the loop coil detector, floating vehicle and adaptive Kalman filter in the traffic peak and accident conditions. The results show that the adaptive Kalman filtering algorithm combines the data from the loop coil detector and the floating car, the predicted value is closer to the measured value and the prediction accuracy is higher.