群体仿真中基于决策树的路径自动评价研究

摘要

针对群体仿真中大量粒子的运动路径评价中存在评价速度慢、准确率低、缺乏依据特征等不足，提出基于决策树算法的路径自动评价模型。在群体粒子自动聚集过程中，通过分析，提取出影响其运动行为的属性，应用于决策树算法中进行路径好坏的判断评价。仿真结果表明，该算法的评价模型能够有效地提高评价速度、建立评价依据并提高评价的准确率。%Existing path evaluation methods of large quantities of particles in the crowd simulation have slow evaluation,low ac-curacy rate and lack of effective features.A automatic route evaluation model based on decision tree was proposed.In the process of automatic aggregation of group of particle,decision tree algorithm was used to determine whether a path was good or bad by analyzing and the attributes that affected their exercise behavior were extracted.The simulation results show that the proposed evaluation model based on the decision tree algorithm can effectively improve the speed of evaluation,establish the evaluation basis and improve the evaluation accuracy rate.