基于多种优化算法的交通事故精准化重建与颅脑损伤风险评估

目的 探究不同优化算法对精准化重建交通事故的影响。方法 应用非支配排序遗传算法(non-dominated sorting genetic algorithm-II, NSGA-II)、邻域培养遗传算法(neighborhood cultivation genetic algorithm, NCGA)以及多目标粒子群优化算法(multi-objective particle swarm optimization, MOPSO)对1例真实案件的多刚体动力学重建进行优化,研究不同优化算法对收敛速度与最优近似解的影响,并将优化所得的最优初始碰撞参数作为有限元模拟的边界条件,将仿真得到的颅脑损伤预测结果与实际损伤进行比较。结果 NCGA算法在优化过程中收敛速度更快,结果更优,最优近似解重建所得行人-车辆碰撞的运动学响应与监控视频相符,颅脑损伤预测情况与尸体检验基本一致。结论 通过优化算法和多刚体、有限元方法结合可以完成交通事故精确化重建,减少人为因素的影响。     

基于有限元法分析钝力性脾损伤的机制

Objective To explore the biomechanical mechanism of blunt spleen injury based on finite element analysis. Methods A fist finite element model was used to simulate the impact at 4-8 m/s in the spleen area of THUMS4.0 human body model from the front of the left costal area, the left anterior axillary line and the rear scapular line. The strain distribution and damage of the spleen under different conditions were observed. The simulation results were compared with the clinical cases of spleen rupture to analyze the mechanism of spleen injury. Results The damage location and strain distribution of the spleen could vary under different conditions. Due to the special anatomical location of the spleen, a blunt impact at the speed of 4-8 m/s on the front side did not easily cause spleen injury, and the strain was distributed mainly in the front of the spleen and the spleen hilum; a similar blunt impact on the left side was likely to cause spleen diaphragmatic surface injury, the splenic visceral surface could be injured by the compression of the medial tissue and organs and the traction of the splenic pedicle, and the strain was distributed in the spleen diaphragmatic and visceral surfaces; an impact on the back side was likely to cause injuries in the posterior portion and hilum of the spleen, and the strain was mainly concentrated in the injured area. Conclusion Blunt spleen injuries caused by punches on the abdomen are mostly caused by direct impact on the ribs, the compression by the surrounding tissues and organs and the traction by the spleen pedicle.