Dynamic Modelling And Stability Analysis Of Traffic Flow With Consideration Of Driving Behavior Characteristics

Recently,the traffic problem has become the main bottleneck to restrict the sustainable development of the city.There are many factors may cause the traffic jam,especially,the driver's driving behavior have a great impact on traffic flow.Therefore,the dynamic modelling and complexity research of traffic flow with consideration of the driving behavior has important values on engineering application as well as realistic significance.Based on the microscopic and macroscopic traffic models,we focus on the driver's driving behavior characteristics such as conservative and aggressive driving,memory effect and honk effect,several improved traffic models are presented,and the corresponding theoretical analysis and numerical simulation are performed.Moreover,the stability and the nonlinear density wave in traffic flows are also investigated.All of these studies can provide certain theoretical guidances for the management and control of traffic flow.The main contents of the thesis are as follows:I.An extended cellular automata model is proposed by considering conser-vative and aggressive driving behaviors.The influence of two types of driving behaviors on the probability of traffic accident are studied,and the microscopic mechanism of traffic accident is also discussed in detail.Based on NaSch and FI cellular automaton traffic flow models,a novel cellular au-tomata model is established by considering conservative and aggressive driving behavior.The influences of the proportion of aggressive drivers,overtaking probability and random braking probability on traffic flow and traffic accident are investigated by numerical sim-ulation,and the typical spatial-temporal dynamics characteristics are also explored.The results show that the driver's aggressive driving behavior is the main factor causing traffic accidents frequently;the higher proportion of the aggressive drivers will induce more traffic accidents;The proportion of rear-end accidents is higher than overtaking accidents in the total accidents.As the density is higher,more conservative driving behaviors can reduce the occurrence of traffic accidents.Furthermore,the processing time of traffic accident become longer,more serious traffic jams may be induced,which will bring significant de-cline in road capacity.The above research can provide some theoretical suggestions for the standard of driver's safety behavior and emergency response planning of traffic accident.II.Three improved car-following models are constructed to simulate traffic flow by considering different memory effects in the process of vehicle move-ment.The influence of different memory effects on the stability of traffic flow is investigated.Moreover,the mKdV equation is obtained by using reduced perturbation method to describe the traffic density wave in unstable region,and the nonlinear characteristics of traffic flow are also discussed.(1)Complex environment(e.g.rainy weather)will interfere with the driver's field of vision,and lead to driver' s memory delay.Motivated by this reason,an extended car-following model is proposed based on FVD model by introducing memory item.Through linear and nonlinear analysis theory,the stability condition and the mKdV equation of this model is obtained,respectively.The results show that the memory effect can lead to the instability of traffic system,and induce to the traffic jam.However,the traffic congestions can be obviously suppressed with seasonable weather due to the decrease of the driver' s memory time.The numerical results are in good agreement with the theoretical analysis.(2)In the real-time traffic state,an extended car-following model is established by considering the transient memory effect of headway and velocity.Theoretical analysis and numerical simulation show that the delay behavior of velocity perception can improve the stability of traffic system and alleviate the traffic congestion effectively.But the delay behavior of headway perception can induce the traffic jam and lead to the occurrence of stop-and-go wave.Therefore,if the driver can strengthen the judgment of headway between the leading vehicle and the following one,and reduce the memory delay caused by the change of headway,the traffic jam can be alleviated effectively.(3)Based on FVD model,an improved car-following mode is presented by considering the driver's continuous memory effect in the process of vehicle movement.Through the stability analysis of this model,it can be found when the continuous memory effect is strengthen,the critical value of the sensitivity will increase and the unstable region will be apparently enlarged.That is to say,traffic flow becomes more unstable.In addition,the numerical simulation results also show that stronger continuous memory effect may lead higher velocity fluctuation,and the traffic congestion is more difficult to dissipate.Through studies of the three memory effects,it is found that the memory delay ef-fect can make the traffic flow instability and induce the traffic jam.In fact,the driver's memory effect is an objective phenomenon,and unmanned driving technology relies on the high real-time sensors and fast algorithms,which can overcome the driver's memory delay.Therefore,our studies can provide a theoretical basis for the necessity of the study about smart cars(e.g.unmanned vehicles).?.Based on the application of intelligent transportation system(ITS),an extended car-following model is proposed to simulate traffic flow by considering average headway of preceding vehicles group.The influence of the dynamic change of average headway on the stability of traffic flow is explored,According to the motion information of many preceding vehicles in an environment of ITS,an extended car-following model is established by considering the average headway of preceding vehicles group.The average headway effect on traffic stability is analyzed through the linear stability theory and nonlinear reductive perturbation method.The neu-tral stability curves depicted in the headway-sensitivity space reveal that the new model has a larger the stable region than that in FVD model.Nonlinear analysis shows that the propagating features of traffic density waves can be illustrated by the kink antikink soliton of the mKdV equation.Numerical simulation is also carried out to check the theoretical analysis.All these results indicate that the average headway effect plays an important role in improving the stability of traffic flow.The above research can provide some theoret-ical basis for the planning of ITS and the formulation of intelligent vehicle driving strategy.?.From the microscopic and macroscopic viewpoint,an extended car-following model and an improved lattice hydrodynamic model were proposed to simulate traffic flow by considering the honk effect.The influence of the driver's honk behavior on the stability of traffic flow is studied.(1)When a vehicle hinders its following vehicle from moving at its current velocity,the following vehicle may honk its horn.Once the preceding driver hears the horn,the driver will probably change lane or accelerate according to the traffic state at that time.Based on FVD model,an improved car-following model is established by considering honk effect.The stability condition of this model is obtained by using the linear stability analysis.Theoretical results show that the honk effect plays an important role in improving the stabilization of traffic system and reducing the occurrence of traffic congestion.By applying the reductive perturbation method,the mKdV equation near the critical point is derived to describe the evolution properties of traffic density wave(e.g.formation,propagation and dissipation of traffic jam).Furthermore,the numerical simulation is consistent with the theoretical analysis.(2)Based on Nagatani's model,an extended lattice hydrodynamic model accounting for the honk effect is presented.The results of linear and nonlinear analysis show that the honk effect always plays an important role in stabilizing traffic flow and suppressing the traffic jam.The above studies show that the traffic jam can be suppressed efficiently form micro-scopic and macroscopic viewpoint.Therefore,in the precondition of avoiding traffic noise pollution,the reasonable planning of the honk area can effectively alleviate the traffic jam and improve the efficiency of transportation in the city.The final chapter of this thesis is devoted to an analysis and prospect of further study of the traffic flow.