Traffic Signal Networks Control Optimize With PSO Algorithm

Dynamic traffic control network has been a key and difficult problem for urban traffic.When the number of vehicles in the city is outgrowing,it causes great pressure on the traffic and surroundings of the city.Moreover,the automobiles' exhaust emission directly contributes to global warming.Traffic congestion has been a serious issue in all big cities around the world.How to control the traffic signal is a problem which needs effective strategy to solve.Many researchers have been working on the optimization of traffic signal adjustment in literature.In this paper,we alleviate this problem by dynamically adjusting traffic signal plans for each node in the traffic network.This paper is mainly engaged in signal control optimization on traffic network to reduce traffic delay and vehicles' stop times,so that to reduce fuel consumption and decrease the emissions.This study can not only reduce the urban traffic congestion,but also mitigate the adverse impact of vehicles on the environment.For traffic signal network optimization problem,we adopt a extended cell transmission model and introduce the serial particle swarm algorithm(TSCOPSA).The main content of the work is as follows:(1)To model the dynamic traffic network,we use the expanded cell transmission model(CTM,Cell Transmission Model).The merge model and the diverge model of basic CTM model is extended.To meet the needs of modern urban dynamic traffic network,we establish the crossroads model.And we describe how to introduce the time interval of traffic signal lights into CTM to provide the specific traffic signal control.Finally,on general AIMSUN and CTMSIM experimental platform,we evaluate the effectiveness of the proposed extended CTM.The experimental results show that the extended CTM is very effective and feasible for dynamic traffic network modeling.(2)For the optimization of signal control in dynamic traffic network,we use serial particle swarm algorithm(TSCOPSA)to find the feasible solution.In dynamic traffic model,the specific the TSCOPSA we adopt is different from the basic particle swarm algorithm.We define the signal optimization fitness function as the vehicle delay,and analyze the impact of traffic signals parameter setting.Empirical scene simulation results show that,compared with Sequential Genetic Algorithm,TSCOPSA achieves better results on delay,fuel consumption,and performance of algorithm itself,respectively,with 3.96%,0.42% and 10.13% of the average improvement;simultaneously,it can reach up to 7.25%,1.03% and 18.2% on the three indicators.Compared with the PPSO on delay,fuel consumption and performance of algorithm,TSCOPSA respectively improve 3.17%,0.22% and 2.03% of the average rate;simultaneously,it can reach up to 5.45%,0.85% and 6.06% on those three indicators.