并行路网结构下的浮动车信息处理模型

FCD based information processing model in parallel road network

受GPS定位精度的限制,浮动车在主辅路并行路网结构下进行地图匹配和路径推测存在困难,直接影响路况信息获取的准确性.在现有浮动车信息处理框架基础上,运用统计学方法提升地图匹配的精度,并将主辅路径同时作为车辆的行驶轨迹输出,有效降低了路径推测错误的可能性.最后,引入证据推理框架以解决路径选择策略调整所带来的信息不可信和信息冲突问题.对比实验表明,改进系统在主路和辅路上的路况准确性分别有14.26%和9.46%的提升.

Limited by the poor GPS locating precision, it is quite difficult to implement map matching and path deriving procedures in the parallel road network by FCD （floating car data） technology, which affects the traffic information accuracy directly. On the basis of existing FCD information processing framework, some approaches were applied to improve map matching performance. And then, the freeway and service road paths were both viewed as vehicle trajectories, so as to reduce the probability of inaccurate path deriving effectively. Finally, the evidential reasoning framework was introduced to address the information unlikelihood and information conflict issues after the path selecting strategy adjustment. Contrast experiment shows that the traffic information accuracy on freeway and service road was improved about 14.26% and 9.46% respectively after employing our modified model.