基于颗粒流离散元方法的行人动力学模型及仿真研究

摘要

Discrete element method is an efficient mechanical tool to analyze and solve complex discrete particle system. It has advantages on improving the solving efficiency of pedestrian dynamic model and depicting mechanical relation among pedestrians. Soft ball contact models are used to model the touching and squeezing behavior of pedestrians, and social force models are used to model subjective motive force as an external input. Based on these, pedestrian movement models are constructed based on the theory of particle flow and implemented on the discrete element simulation platform EDEM. The validity of these models is verified by comparisons between simulation data and field date under different pedestrian densities. The high efficiency of this algorithm is proved by calculation time comparison between different simulation platforms. The study results show that: the simulated pedestrian flow behavior under different densities are in line with field observations, as so as the average speeds of pedestrians, and the simulation model has high calculation efficiency. Therefore, the pedestrian simulation models based on particle flow theory and discrete element method can be realizable and need deeper research.%离散元方法是分析和求解复杂离散颗粒系统问题的一种高效力学数值模拟方法，对于提高目前行人动力学模型中的求解效率，细致刻画行人运动中的力学关系具有优势。本文利用颗粒接触理论中的软球模型描述行人的接触及挤压行为，以现有社会力模型中的主观运动力作为外部输入，构建了基于颗粒流理论的行人运动模型，并在离散元EDEM仿真平台上实现。通过对不同密度条件下行人运动行为及平均速度的仿真及统计，对比实测数据，验证了模型的有效性；通过对比不同仿真平台的计算速度，验证了算法的高效性。研究结果表明：模拟得到的行人流在不同密度下的行为符合现场观测，模拟所得行人平均速度数据与实测数据基本吻合，所建模型具有较高计算效率。因此，基于颗粒流理论的离散元方法能够实现对行人运动的模拟，且具有进一步的研究价值。