文献类型：期刊文献

中文题名：基于碰撞预测的障碍物社会力修正模型

英文题名：A Revision Model for Obstacle Social Force Computation Based on Collision Prediction

作者：于萍[1]

第一作者：于萍

机构：[1]新乡学院管理学院

第一机构：新乡学院管理学院

年份：2017

卷号：34

期号：12

起止页码：39-42

中文期刊名：新乡学院学报

外文期刊名：Journal of Xinxiang University

基金：国家自然科学基金项目(41571393)

语种：中文

中文关键词：障碍物作用力修正;碰撞预侧;社会力模型;反弹

外文关键词：revision of obstacle force; collision prediction; social force model; rebounding; passing-through

摘要：针对经典社会力模型中行人与障碍物碰撞后,可能出现的"反弹"现象,提出基于碰撞预测的障碍物作用力模型。模型假设障碍物和行人均为刚性介质,在障碍物作用力与行人的运动状态(如速度、加速度)以及相对位置之间建立关系,通过碰撞预测,使障碍物产生与行人运动状态相当的作用力。实验证明,模型可有效避免"反弹"现象的发生,使社会力模型对行人行为的模拟更符合实际。
A collision prediction model was proposed to eliminate the possible phenomena as “rebounding-from” or “passing- through” the obstacles after pedestrians’ colliding with the obstacles in crowding simulation based on traditional social force model. The collision prediction model supposed that the obstacles and the pedestrians were all rigid, and then bui lt relationships between obstacle forces and pedestrian^ motion parameters （like velocity, acceleration）, and the comparative positions. Through collision prediction the obstacle forces could achieve dynamic consistent force with the state of the pedestrian. Experiment results showed that the proposed model could effectively avoid the problems of “rebounding-from” or “passing-through” the obstacles,which made the crowd simulation more realistic.