Modeling And Control Of Traffic Flow Under Complex Traffic Scenarios

With the rapid development of social economy,traffic problems such as traffic congestion have drawn considerable attention due to the explosive growth of vehicle population.To address the above-mentioned issues,a promising solution is to make scientific traffic policies based on the accurate traffic flow models.Hence,there is an emergent research need to caputre the characteristics of traffic flow such that the scientific traffic policies can be made accordingly.Recently,connected vehicle(CV)technologies such as vehicle-to-vehicle/vehicle-to-infrastructure(V2V/V2I)communications are able to provide advanced means to perceive the comprehensive information of vehicles and then characterize the behavior of traffic flow accurately.Consequently,it is of theoretical and practical significance for congestion mitigation based on the modeling and control of traffic flow under the complex scenarios via the CV technologies.In this context,this thesis focuses on cognizing the behavior of traffic flow under complex traffic scenarios as well as designing the reasonable control strategy from the microscopic and macroscopic perspectives,respectively.In particular,on one hand,some extend traffic flow models are proposed by considering the effects of different traffic scenarios at the microscopic level.In addition,the steady-state and dynamic performance of the proposed models are analyzed,respectively,to capture the characteristics of traffic flow.On the other hand,a delay feedback control strategy is developed based on the lattice model,at the macroscopic level,to smoothen traffic flow and improve the road capacity.The main work and contributions of this thesis are summarized as follows.(1)Regarding non-lane-discipline-based scenario,a new microscopic traffic flow model is proposed by considering the effects of two-sided lateral gaps.Also,the effects of different lateral gaps on energy consumption of electric vehicles are investigated accordingly.Considering the non-lane-discipline scenario in practice,the lateral gaps between the following vehicle and the two-sided preceding vehicles are characterized by defining a specific variable.Consequently,a two-sided full velocity difference model(TSFVD)is proposed by incorporating the effects of lateral gaps such as to extend the lane-disciplinebased scenario to non-lane-discipline-based scenario.Stability analysis of the proposed TSFVD model is performed using the perturbation method.Results from theoretical analysis and numerical experiments show that the steady-state and dynamic performance of the TSFVD model is better than those of the existing models with respect to the stable region and space headway.In addition,an energy consumption model including the energy loss and energy recuperation is presented.And,the relationship between the traffic flow model and energy consumption model is explored based on the velocity and acceleration profiles.Finally,the energy consumptions of electric vehicles are systematically investigated under the scenarios of no lateral gap,one-sided lateral gap,and two-sided lateral gaps.Results from numerical experiments demonstrate that the energy consumption of electric vehicle traffic stream will increase due to the lateral gap.(2)Regarding V2 V communication scenario,a microscopic traffic flow model is developed by incorporating the effects of vehicle dynamics.The behavior of traffic flow is affected by the state of individual vehicle and also can impact the movement of individual vehicle.Under V2 V communication scenario,various information can be exchanged between vehicles by leveraging the capability of V2 V connection.In order to reveal the relationship between the internal factor and external state of vehicle,vehicle dynamics information(e.g.the opening angle of electronic throttle)is further introduced.Consequently,a throttle-based full velocity difference(T-FVD)model is proposed incorporating the effect of the opening angle of electronic throttle.In addition,stability analysis of the proposed T-FVD model is performed using the perturbation method.Results from theoretical analysis and numerical experiments indicate that the steady-state and dynamic performance of the T-FVD model are better than those of the existing models with respect to the stable region and position,velocity,acceleration/deceleration as well as space headway profiles.(3)Regarding V2 I communication scenario,a microscopic traffic flow model is proposed by considering the effects of both roadside device communication and lateral gap.The roadside device can share the information of vehicle state and road structure with the individual vehicle within the communication range through the V2 I communication,which has an impact on the state of individual vehicle and then potentially affects the behavior of traffic flow accordingly.In order to explore the effect of roadside device on the behavior of traffic flow under the non-lane-discipline-based scenario,based on the definitions of the lateral gap between the following and preceding vehicles as well as the distance between the following vehicle and the roadside device,a non-lane-disciplineroadside-based car-following(NLDR)model is proposed by incorporating the intensity of V2 I communication signal.Then,stability analysis of the proposed NLDR model is performed using the perturbation method.Results from theoretical analysis and numerical experiments demonstrate the steady-state and dynamic performance of the NLDR model are better than those of the existing models with respect to velocity and acceleration/deceleration profiles.In addition,energy consumption is used to measure the evaluation index,results from numerical experiments show that the energy consumption of electric vehicle traffic stream will reduce due to the V2 I communication.(4)Based on the macroscopic lattice model of traffic flow,a delay feedback control strategy is developed by considering the effect of density change rate.The aforementioned sections(1),(2),and(3)focus on modeling the traffic flow under complex traffic scenarios to capture the inherent characteristics of traffic flow such as to provide theoretical basis for congestion mitigation and energy consumption reduction.Considering the backward propagation and delay in traffic flow,a lattice model based delay feedback control strategy is developed by taking the difference of density change rate as the feedback signal.In addition,stability analysis is performed using the small gains theorem to obtain the stability condition.Results from numerical experiments,including single perturbation and multiple perturbations,illustrate that the designed control strategy can reject the propagation of perturbations effectively and smoothen the traffic flow steadily so as to mitigate traffic congestion.