基于参数灵敏度分析的交通仿真模型校正研究  

Research on traffic simulation model correction based onparameter sensitivity analysis

作　　者：张建旭[1] 王雪芹 Zhang Jianxu;Wang Xueqin(College of Traffic&Transportation,Chongqing Jiaotong University,Chongqing 400074,China)

机构地区：[1]重庆交通大学交通运输学院,重庆400074

出　　处：《计算机应用研究》2022年第6期1790-1795,共6页Application Research of Computers

基　　金：科技助力经济2020重点专项(SQ2020YFF0418521);中央引导地方科技发展专项(CSTC2020JSCX-DXWTB0003)。

摘　　要：为了使交通仿真模型校正工作能够高效开展,提出了以参数灵敏度分析为基础的模型校正框架。通过灵敏度分析确定影响模型精度的关键参数,以简化模型;对关键参数进行标定,以校正模型。以城市快速路交织区为仿真案例,以跟车模型和换道模型为研究对象;首先进行了大量仿真实验,分析不同车流量水平下模型参数的取值特征;据此制定模型参数的区间划分规则和交叉组合规则,从而对LH-OAT算法和遗传算法(GA)进行改进;然后应用改进LH-OAT算法(ILH-OAT)对模型参数进行灵敏度分析,再应用GA对关键参数进行标定;最后依据校验指标对仿真结果进行误差分析。结果表明ILH-OAT和GA相结合,不仅简化了仿真模型,降低了仿真运行成本,仿真效果也更加接近真实的道路交通运行情况。In order to carry out the correction work of traffic simulation model more efficiently,this paper proposed a model correction framework based on parameter sensitivity analysis.The method simplified the model by exploring the key parameters that affected the model accuracy through sensitivity analysis,and corrected the model with the usage of the calibrated key parameters.To justify this,the paper analyzed the car following model and lane changing model under the simulation condition of urban expressway weaving area.Firstly,it obtained the value of model parameters under different traffic flow levels through a large number of simulation experiments.Secondly,this paper formulated the interval division rules and cross combination rules of model parameters to improve LH-OAT algorithm and genetic algorithm(GA).Then,it used the improved LH-OAT algorithm(ILH-OAT)and GA algorithm to analyze the sensitivity of the model parameters and used GA to calibrate the key parameters.Finally,this paper checked the error of the simulation results according to the verification index.The results show that the combination of ILH-OAT and GA simplifies the simulation model and reduces the simulation operation cost.Besides,it also adds verisimilitude to the simulation on reflecting real traffic operation.

关 键 词：参数灵敏度分析 关键参数标定 ILH-OAT算法 遗传算法 

分 类 号：TP399[自动化与计算机技术—计算机应用技术]