The crash between two-wheelers and passenger vehicle causes a higher fatality rate and many casualties for the two- wheelers riders. Both two-wheelers and pedestrians are considered as VRU. Compared to pedestrians, two-wheelers are featured with fast moving speed and uncertain driving route, which poses a challenge to the optimal control for the autonomous emergency braking (AEB) system. This paper firstly screened 216 cases of frontal collision accidents between passenger vehicles and two-wheelers from the database of National Automobile Accident In- depth Investigation System (NAIS) in China, extracted the static and dynamic variables related to the pre-crash scenarios reconstruction in each case. This paper extracted four typical pre-crash scenarios between two-wheelers and passenger vehicles from 216 accident scenarios through clustering analysis and chi-square test, reconstructed and simulated typical pre-crash scenarios by using PreScan software and completed matching, optimization and analysis on the field of view (FoV), braking trigger width (w), time to collision (TTC) of AEB system to obtain the boundary parameter conditions of the AEB system to avoid crash or greatly reduce the collision speed, providing a reference for the development of AEB systems applicable to China's road traffic scenarios. The research method used in this paper is applicable to the reconstruction and simulation analysis of pre-crash scenarios for passenger vehicles and pedestrians as well as the parameter optimization of other ADAS. In addition, the research method used in this paper also provide a technical solution for the design, test and evaluation of the automatic driving function based on typical scenarios.