微观交通仿真模型的全微观参数校正

Identity Microscopic Parameters Calibration of Microscopic Traffic Simulation Models

参数校正是微观交通仿真模型应用前的必要环节。现有的微观参数校正方法以宏观实测交通数据作为校正数据,因而工作量大、耗时长、参数调整困难且校正结果精度有限。针对这些问题,提出了一种利用微观实测交通数据对微观仿真模型进行参数校正的方法——全微观参数校正,提出了相应的校正流程,并给出了校正的目标函数,求解方法,以及数据处理等核心步骤的实现。在此基础上,利用NGSIM标准微观实测数据对经典的微观模型进行了全微观参数校正实验,通过分析实验结果,表明了全微观参数校正的合理性和有效性。

Parameters Calibration is indispensable before applying Microscopic Traffic Simulation models.The existed ways of Microscopic Parameters Calibration（MPC） are usually performed by utilizing the observed macroscopic traffic data as the calibration data.Therefore,it is time-consuming and brings heavy workload,difficult parameters-tuning,dissatisfied calibration results.Facing to these problems,an approach of utilizing the observed microscopic traffic data to perform MPC was proposed,which could be called Identity MPC（IMPC）.The detail of calibration flow was proposed,and some key steps were described as well,such as the establishment of objection function,the problem solving and the data processing.Based on Next Generation SIMulation（NGSIM） observed microscopic standard traffic data,the well-known microscopic traffic models were calibrated by IMPC,and the experiments results prove the validity of IMPC.