基于改进微粒群算法的单点信控交叉口配时优化

摘要

With the development of urbAn trAffic,the signAl timing optimizAtion of intersection hAs the extremely vi-tAl role to relieve the trAffic congestion of urbAn roAd And enhAnce cApAcity of urbAn roAd. TAking the typicAl four phAses of single intersection As exAmple,the Article tAkes the totAl delAy,number of stops And cApAcity As the opti-mizAtion objective in eAch phAse. According to the reAl-time trAffic dAtA,the corresponding weight coefficient hAs been Adjusted. BecAuse pArticle swArm Algorithm to solve constrAined optimizAtion problems is good for function op-timizAtion problems,the signAl timing scheme wAs obtAined bAsed on the model which wAs solved by using A con-strAined optimizAtion problem of the pArticle swArm Algorithm. The simulAtion results show:compAred with the trAdi-tionAl Webster Algorithm,new signAl timing scheme is effective. It is more suitAble for single intersection optimizA-tion control,And it provides A better line control And pAnel control of trAffic for further AnAlysis.%交叉口的信号控制，对减轻城市道路的交通拥挤，提高城市道路通行能力有极其重要的作用。以典型的四相位单点控制交叉口为例，选取每个相位进道口上的总延误时间、车辆的停车次数和道路的通行能力作为优化目标。由于求解约束优化问题的微粒群算法有利于函数型优化问题，所以利用该算法对模型进行求解，得到新的信号配时方案。仿真结果表明，与传统的 Webster 算法进行比较，由求解约束优化问题的微粒群算法所得到的信号配时方案是更优，更适合于单交叉口进行信号优化控制，为进一步分析研究城市交通线控、面控提供更好的方法。