基于视频信息的高精度事故重建方法研究

Research on High Precision Accident Reconstruction Method Based on Video Information

为提高事故重建精度,本文中采用多体系统与有限元相结合的方法对两起具有清晰视频信息的VRU事故案例进行重建,验证并提出了一种高精度事故重建方法。首先,采用直接线性变换理论(DLT)对视频信息分析并获取碰撞车速;然后,通过PC-Crash与MADYMO耦合方法精确重建VRU碰撞运动学响应;最后,采用THUMS(Ver 4.0.2)人体有限元模型对VRU颅内运动学和动力学参数深度分析。结果表明:重建所得的车辆-VRU碰撞部位、VRU碰撞过程的运动学响应、旋转角度、落地姿态与着地部位与视频信息完全一致;VRU与车辆最终位置绝对误差小于7%;VRU头部损伤部位及损伤严重度与伤情报告完全吻合。

In order to improve the accuracy of accident reconstruction,the method of combination of multi-body system and finite element is used in this paper for reconstruction of two VRUs accident cases with clear video information and a high-precision accident reconstruction method is proposed and verified.Firstly,the video information is analyzed and vehicle collision speed is obtained based on the direct linear transformation(DLT)theory;then the VRUs kinematic response is accurately reconstructed by using the PC-Crash and MADYMO coupling method;finally,the intracranial kinematics and kinetic parameters are deeply analyzed based on the human body finite element model THUMS(Ver 4.0.2).The results show that the reconstructed vehicle VRU collision position,the kinematic response,rotation angle,landing posture and landing body of the collision process are completely consistent with the video information.The absolute error between VRU and vehicle final position is less than 7%,and the head injury position and injury severity of VRU are consistent with the injury report.