基于Q-paramics的微观交通仿真模型参数校正  

Parameter-correction of Microscopic Traffic Simulation Model Based on Qparamics

作　　者：成卫[1] 袁满荣 陈辉[1] 

机构地区：[1]昆明理工大学云南省(高校)智能交通系统工程研究中心,云南昆明650093 [2]昆明市公安局交通警察支队,云南昆明650023

出　　处：《系统工程》2013年第2期42-46,共5页Systems Engineering

基　　金：云南省科技创新基金资助项目(2008QA022)

摘　　要：为使交通仿真模型能够更好地运用于各种实际交通运行方案的测试中,需要对交通仿真模型进行参数标定。在微观交通仿真模型的基础上,提出微观交通参数及动态路径选择的校正设计流程;通过对仿真模型的敏感性分析,选取对参数校正影响较大的5个参数进行研究,然后利用遗传算法进行参数校正,并采用试错法将路径校正纳入整个参数校正流程当中。构建以昆明市一二一大街、学府路为主的区域路网并在Q-paramics软件中仿真模拟,最后选取交叉概率为0.5、变异概率为0.1的仿真因子,搜索出最优参数组合。仿真结果显示,相对误差基本控制在10%左右证明了校正结果的合理性及具体实现方法的有效性。To make a better use of the traffic simulation model to test kinds of real traffic running programs, it is necessary to calibrate the parameters of the traffic simulation models. Based on the microscopic traffic simulation model, the designing process of the correction for microscopic traffic parameters and dynamic, path selection was put forward; through the sensitivity analysis of the simulation model, the five parameters more influenced to parameter-correction were studied and then the genetic algorithms was used in parameter calibration and the path correction put into the whole parameter- correction process by adopting the route-correction method. The area network mainly consist of 121 street and xuefu road was constructed, which was simulated in Q-paramics, and it was finally found out the optimal parameter combination used the simulation factors, which was selected that the crossover probability was 0. 5, the mutation probability 0. 1. The simulation result shows that keeping the relative error at 10% qualifies the rationality of the correction result and the validity of the specific implementation method.

关 键 词：智能交通系统 参数校正 遗传算法 城市路网 

分 类 号：U491[交通运输工程—交通运输规划与管理]