我国人车混行现象较为严重，行人在交通事故中占有较高的伤亡比例，行人碰撞安全问题已经成为汽车被动安全领域的热点问题。本文基于真实人车事故数据，建立人车碰撞模型，研究了人车事故中初始碰撞参数、行人碰撞后运动及损伤参数的分布规律，系统研究了初始碰撞参数对行人碰撞后运动及损伤参数的影响规律，最后定量预测和对比研究了主动、被动及主被动一体化保护措施的行人防护效率。基于Madymo平台建立了适用于人车碰撞研究的车辆多面体模型，并通过相关试验数据验证了模型的有效性；针对弗吉尼亚大学初步建立的多面体人体模型进行二次开发，采用最优化方法对模型颈部力学特性进行了优化改进，同时对人体模型生物逼真度进行了验证。进一步建立了人车碰撞模型，并通过再现真实人车事故对模型有效性进行验证，为后续开展蒙特卡洛仿真试验奠定基础，同时模型也可为人车事故再现的具体实践奠定重要基础。基于清华大学道路交通事故数据库中典型人车事故数据，通过统计分析方法得到了初始碰撞参数的统计分布特征，以此作为生成初始碰撞参数随机数的依据；在此基础上，采用蒙特卡洛方法开展了2660次人车碰撞仿真试验，系统研究了行人碰撞后运动及损伤参数的统计分布规律，建立了行人碰撞后运动及损伤参数的概率分布模型，该模型可为发动机罩和风挡玻璃的吸能特性研究提供源于真实事故数据的头部碰撞工况，也为行人安全气囊、弹起式发动机罩的开发设计提供重要的数据支持。通过参数化方法和统计学非参数检验方法，系统研究了汽车碰撞速度与其余初始碰撞参数对行人运动及损伤参数的耦合影响，建立了行人运动及损伤参数的置信水平为0.90的置信区间模型及行人损伤风险预测模型；并最终建立了能够描述碰撞初始参数与行人损伤指标之间关系的多元回归模型。上述模型揭示了人车碰撞过程中行人运动及损伤参数在初始碰撞参数影响下的变化规律，可为人车事故鉴定工作提供理论参考。通过引入主动、被动及主被动一体化保护措施，进一步开展人车碰撞仿真试验，预测并对比研究了我国道路交通状况下主动、被动及主被动一体化措施的行人防护效率，以及碰撞速度对主动、被动及主被动一体化保护措施防护效率的影响，为针对我国道路交通状况开发主被动一体化保护系统的可行性提供了理论支持。

Vehicle-pedestrian mixed traffic condition is serious in China, pedestrians always occupy a higher proportion of casualties in traffic accidents, so pedestrian collision safety has become a hot issue in the field of passive safety. In this paper, based on the data of real vehicle-pedestrian accidents, the distribution of initial collision parameters as well as the pedestrian’s kinematics and injury parameters after the collision are studied. The influence of initial collision parameters on the movement and injury parameters after collision is systematically studied.Based on the Madymo platform, a facet vehicle model suitable for vehicle-pedestrian collision research is established, which is validated by the relevant experimental data. Secondary development is conducted based on the facet human model originally established by the University of Virginia, an optimization method is applied to improve the mechanical properties of the neck model, and the bio-fidelity of the human model is validated. A simulation model of vehicle-pedestrian collision is further established, which is validated by the reconstruction of a real vehicle-pedestrian accident, the simulation model lays the foundation for the follow-up Monte-Carlo simulations. Meanwhile, the model can also lay an important foundation for the practice of vehicle-pedestrian accident reconstruction.Based on the typical vehicle-pedestrian accident data from the traffic accident database of Tsinghua University, the statistical distribution of the initial collision parameters is obtained by statistical analysis, which will be used as the basis for generating random numbers of the initial collision parameters. On this basis, a total of 2660 vehicle-pedestrian collision tests are carried out using the Monte Carlo method. The statistical distribution of the kinematics and injury parameters after the collisions is systematically studied. The probability distribution model of the kinematics and injury parameters after the collisions is established. The model can be used to provide head collision condition from the real accident data to the research related to energy-absorption characteristics of hood and windshield. It also provides important data support for the development and design of the pedestrian airbag and pop-up hood.By parameterization method and statistical nonparametric test method, the coupling influence of vehicle collision speed and other initial collision parameters on pedestrian motion and injury parameters are systematically studied. A confidence interval model with confidence level 0.90 and pedestrian injury risk prediction model are established. Finally, a multivariate regression model that can describe the relationship between initial parameters of collisions and pedestrian injury indicators is established. The above models reveal the changing rules of pedestrian kinematics and injury parameters under the influence of initial collision parametersin pedestrian-vehicle accidents, which can provide theoretical reference for the identification of pedestrian-vehicle accidents.Active, passive and active-passive integrated protection countermeasures are introduced to further Monte Carlo simulations. The effectiveness of active, passive and integrated active-passive protective countermeasures for pedestrian protection under Chinese traffic condition are predicted and comparative studied, as well as the influence of collision speed to the protection effectiveness of active, passive and active-passive protective measures. It provides theoretical support for the feasibility of active-passive integrated protection system under Chinese traffic conditions.