| Driver is a significant participant in the complex system called "driver-vehicle-road".Their perception,judgment and manipulation will influence the driving condition.The effect may cause traffic congestion or even traffic accidents.Therefore,the driver factor is significant for traffic operation.The previous traffic simulation model fails to display the difference of the driver’s mental state.Hence,by taking the driver’s attention dispersion as an example,this paper separately based on the traffic simulation method of STCA cellular automation and the MultiAgent to explore the approach to build driver’s model under the situation which the driver’s psychological state changes.Meanwhile,this paper also carries out simulation on the traffic flow operation state when a certain proportion of the drivers distracted.To study the driver’s state change and its impact on the traffic flow under the sub-task condition,this paper firstly makes a cognitive structure on the driver’s distraction based on the ACT-R cognitive architecture and the Distract-R platform.Then obtains the ratio of time consumption and distraction when executing four different sub-tasks.The data are used to build the driver’s distraction database.Next,based on the STCA cellular automaton traffic flow model,a new traffic flow model which take the effect of sub-tasks into consideration was built.This model has corrected the deceleration rule of original cellular automaton and obtain part of the model parameters by calling the driver’s distraction database.Further more,the traffic flow simulation experiment was carried out in Matlab to simulate the traffic flow condition when 0,10% and 20% of the drivers perform sub-tasks in driving.The type of sub-tasks was chosen from the above four types at random.In addition,based on Multi-Agent simulation method,this paper established Road Agent,Vehicle Agent and Light Agent successively,and made operation and interaction rules for each agent.Moreover,this paper utilized NetLogo platform to program and built the Multi-Agent traffic simulation system which consider the driver’s distraction and then used this system to carried out simulation experiment.By comparing the results of different experimental,they can verify and supplement each other.The experimental data showed that the sub-task will have a significant impact on the traffic flow: the results of cellular automaton simulation experiments found that when 10% and 20% of the drivers execute sub-tasks,the maximum traffic flow was reduced by 21.4% and 36.2% and the maximum speed was reduced by 11.1% and 22.2%.Meanwhile,with the increase of the proportion of driver who execute sub-tasks,the density which corresponding to the peak of traffic flow and vehicle speed decreased.Based on the Multi-Agent programming method,the results of simulation experiment on road model showed that when 10% and 20% of drivers executing sub-tasks,the maximum traffic flow was reduced by 20.3% and 37.6%,the maximum speed was reduced by approximately 15.6% and 29.3%.In the simulation experiment on road network model,the peak value of vehicle agglomeration increased,and the dissipation was slow down,while the congestion in crossroads was increased.The experimental results showed that the results obtained from the two simulation methods are basically coherent,which meet the classical trend in the previous literature.The results can reflect the changes of the traffic flow to some extent when the driver was distracted.Further more,the experiment also revealed that when some drivers were distracted,the overall state of traffic flow within the road will be significantly influenced.For example,the actual traffic capacity will reduce and the saturated traffic density will decrease. |
