基于改进遗传卡尔曼算法的短时路段行程时间估计  

Estimation of short-time road travel time based on Kalman filter optimized by genetic algorithm

作　　者：林园 LIN Yuan(Urban Development and Planning Institute,Shanghai Tongji Urban Planning and Design Institute Co.,Ltd.,Shanghai 200092,China)

机构地区：[1]上海同济城市规划设计研究院有限公司城市开发规划研究院,上海200092

出　　处：《山东交通学院学报》2020年第1期32-39,53,共9页Journal of Shandong Jiaotong University

基　　金：国家自然基金面上项目资助项目(51678505)。

摘　　要：为进一步提高城市道路行程时间短时预测的准确性,利用城市浮动车全球定位系统(global positioning system,GPS)数据,提出基于遗传算法(genetic algorithm,GA)优化的卡尔曼滤波(kalman filtering,KF)模型估计短时道路行程时间的方法。利用高斯滤波器对异常数据进行有效剔除,基于城市道路地理信息系统(geographic information system,GIS)数据提出道路地图匹配算法,利用卡尔曼滤波器预测匹配道路的行程时间,并通过遗传算法优化卡尔曼滤波的误差参数。实际算例表明:预测路段的行程时间误差均小于0.5 min,基于GA优化的KF算法能有效提高路段行程时间估计的精度。根据预测行程时间对上海市外环路各个路段不同时间的拥堵情况进行预测识别,预测结果有助于交通管理部门及时掌握城市道路运行状态信息。In order to further improve the accuracy of short-time prediction of urban road travel time,and according to the global positioning system(GPS)data of an urban floating vehicle,an optimized Kalman Filter(KF)model based on a genetic algorithm(GA)is proposed for estimating the short-time travel time of a road segment on the basis of existing research.The gaussian filter is used to eliminate the abnormal data effectively.A road map matching algorithm is constructed based on the geographic information system(GIS)information of the urban road.The travel time of matching road is predicted by the Kalman filter,and the error parameters of the Kalman filter are optimized by the genetic algorithm.The actual case shows that the travel time error of the predicted segment is less than 0.5 min,and the KF algorithm based on GA optimization can effectively improve the accuracy of the travel time estimation of the road section.According to the predicted travel time,the congestion of each section of Shanghai outer ring road at different times is predicted,which is helpful for traffic management departments to timely grasp the information of urban road running status.

关 键 词：GA KF 行程时间估计 浮动车GPS数据 

分 类 号：U491.1[交通运输工程—交通运输规划与管理]