基于SOGA的VISSIM仿真模型参数标定方法

发布时间：2018-04-16 14:09

本文选题：智能交通 + 微观交通仿真　；参考：《交通运输系统工程与信息》2017年03期

【摘要】：微观交通仿真模型是对交通系统进行管理、控制和优化的重要试验手段和工具,而微观交通模型的参数标定是确保微观交通仿真模型能真实、直观地反映交通流运行情况的必要前提.针对遗传算法(GA)的不足,提出了基于自适应正交遗传算法(SOGA)的微观交通仿真模型参数标定方法.选取应用较为广泛的VISSIM仿真模型作为基础平台,给出了该优化方法中染色体的编码解码、适应度函数和自适应正交交叉算子的详细设计.最后将算法应用到北京市荣华中路与荣京西街交叉口模型参数标定中,通过与GA算法、正交试验法对比,SOGA算法得到的适应度函数值为19.43,优于其他标定算法的适应度函数值;同时,SOGA算法迭代时间比GA算法少了40.5%,验证了SOGA算法在VISSIM参数标定上的优越性.
[Abstract]:Microscopic traffic simulation model is the management of traffic system, control and optimization of test instruments and tools, and the parameters of microscopic traffic model calibration is to ensure that the microscopic traffic simulation model can truly reflect the necessary premise, traffic flow. According to genetic algorithm (GA) problems, proposed an adaptive orthogonal genetic algorithm (SOGA) calibration method based on microscopic traffic simulation model parameters. Select the VISSIM simulation model is widely used as a basic platform, gives the optimization of chromosome encoding and decoding method, adaptive function and detailed design of adaptive orthogonal crossover operator. Finally, the algorithm is applied to Ronghua Road in Beijing city and Rongjing street intersection model calibration, by comparison with the GA algorithm, the orthogonal test method, SOGA algorithm to get the fitness function value is 19.43, better than the other calibration algorithms of fitness At the same time, the iterative time of the SOGA algorithm is 40.5% less than that of the GA algorithm, which validates the superiority of the SOGA algorithm in the calibration of VISSIM parameters.

【作者单位】：北京交通大学交通运输学院;北京交通大学轨道交通控制与安全国家重点实验室;北京交通大学城市交通信息智能感知与服务工程技术研究中心;
【基金】：国家科技支撑计划课题(2014BAG01B02)~~
【分类号】：U491.1

【参考文献】