中文摘要：

利用车载尾气排放检测系统（PEMS）对捷达轻型车在高速公路人工收费（MTC）和电子收费（ETC）两种方式下的大量尾气排放数据进行了收集和比较．分析结果表明ETC能降低所有污染物的排放，但NOx的降幅却远低于HC和CO的降幅．应用车辆比功率（VSP）模型方法进行了深入的排放预测研究，研究发现[-2，2]是收费站尾气排放预测的重要VSP区间．对[-2，2]进行VSP区间细分后，排放预测的准确度得到了提高．对车辆速度、VSP、NOx累积排放的全面分析解释了NOx降幅低于其他污染物的原因．最后对利用排队长度的排放预测模型进行了研究，发现该模型不适用于MTC方式下的NOx排放预测．

英文摘要：

The real-world vehicle emission and activity data of a light-duty-gasoline vehicle, Jetta, were collected at a freeway toll station using PEMS. The data on both electronic toll collection （ETC） lanes and manual toll collection （MTC） lanes were collected for comparison. Data analysis showed that all emissions have been reduced using ETC. However, the emission reduction of NOx was much less than those of HC and CO. Further, vehicle specific power （VSP） based emission estimation approaches were studied using the collected data, in which some errors in common were identified. For the toll station, the VSP interval [ - 2, 2] was found to be a critical interval in the driving modes as well as in the average emission rates. A finer binning approach was applied, and the accuracy of the emission estimation for the toll station was improved. Subsequently, a comprehensive analysis on speed, VSP, and the accumulated emissions around the tollbooths was conducted to explain why the percentage of NOx reduction was much less than those of HC and CO. Finally, a queue-based emission estimation model was studied for MTC lanes, in which the model was found not suitable for the estimation of NOx emission.