Numerical simulation and analysis of influencing factors of vehicle-pedestrian accidents in untypical contact state

Current numerical analysis of car-pedestrian accidents pays more attention to typical contact state, which means pedestrians collide with the front of vehicle, but less to untypical contact state, which means pedestrians collide with the edge of vehicle. The keystone of this study is to discover appropriate initial parameters of accidents in untypical contact state fast and accurately because numerical simulation usually takes long time. Meanwhile, functions between optimised parameters and initial input parameters in an untypical-contact-state accident are fitted by polynomial regression analysis and has achieved a high degree of accuracy. Simulation of typical case indicates that the combination of orthogonal test and polynomial regression analysis can determine appropriate initial parameters more accurately compared with orthogonal test. Furthermore, the results of numerical simulation and medical identification have achieved a high degree of consistency, which indicates that the reliability and rationality of simulation have been verified.