基于卡尔曼滤波与大数据预测的路径规划研究

Research on path planning based on Kalman filtering and big data prediction

针对近年来严重的交通拥堵现象,提出了卡尔曼滤波和交通流大数据相结合的行程时间预测模型,Dijkstra算法规划最短时间路径方法。首先利用某时刻前3个时段的车辆平均速度数据进行卡尔曼滤波预测同一路段若干个时段后的路段行程时间。然后通过理论最短时间、大数据分析对卡尔曼滤波预测进行优化,得出最佳的预测行程时间。最后根据Dijkstra算法规划出最短时间路径。结果表明设计方法各方面误差指标均优于原模型,交通流大数据与卡尔曼滤波预测相结合的预测方法更加精确有效。

In view of the serious traffic congestion in recent years,a travel time prediction model based on Kalman filter and traffic flow data is proposed.Dijkstra algorithm is used to plan the shortest time path.Firstly,Kalman filter is used to predict the travel time of the same section after several periods using the average vehicle speed data of the first three periods at a certain time.Then the Kalman filter prediction is optimized through theoretical shortest time and large data analysis,and the optimal travel time is obtained.Finally,the shortest path is planned according to Dijkstra algorithm.The results show that the design method is better than the original model in all aspects of error indicators,and the prediction method combined with large traffic flow data and Kalman filter prediction is more accurate and effective.