为了消除跟驰模型用于交通排放测算产生的系统误差，提高微观交通仿真模型与交通排放模型融合应用于交通排放测算的准确性，引入衡量加速度变化率的参数急动度（jerk），提出一种考虑jerk分布特征的Wiedemann跟驰模型优化方法。通过对实际轨迹数据与车辆跟驰数据的拟合分析，建立各加速度区间下的jerk分布，并增设jerk约束对交通仿真模型生成轨迹进行优化。以北京市出租车轨迹数据和跟驰数据为例，测算结果表明：Wiedemann模型优化后，不同速度区间的车辆比功率（Vehicle Specific Power, VSP）分布误差平均降低1.2%，CO₂, CO, THC, NO_X等4种排放物的排放因子平均误差分别降低了16.9%, 118.3%, 27.0%, 20.5%，表明该优化方法能够有效改善原模型中不真实的加速度，降低排放测算误差。

In order to eliminate the systematic error caused by car-following model used in traffic emission measurement, and improve the accuracy of traffic emission measurement from the integration of micro-traffic simulation and traffic emission model, this paper introduced the parameter jerk which was used to measure the acceleration change rate and proposed an optimization method of Wiedemann car-following model considering the characteristics of jerk distribution. By analyzing the field trajectory data and car-following data, the jerk distribution under each acceleration interval was established, and the jerk constraint was added to optimize the trajectory generated by traffic simulation model. Based on the trajectory data and car-following data collected from taxis in Beijing, the results showed that after the optimization of the Wiedemann model, the distribution error of VSP(Vehicle Specific Power) in different speed interval was reduced by 1.2% on average, and the relative errors of emission factors decreased by 16.9%, 118.3%, 27.0% and 20.5% for CO₂, CO, THC and NOx respectively, which means that the optimization method can effectively correct the unrealistic acceleration from the original model and reduce the emission measurement error.