| Driving behavior has close relationship with traffic congestion, and the formationprobability of traffic congestion can be reduced by optimizing the driving behavior.Traffic flow theory is a powerful tool to describe the relationship between drivingbehavior and traffic congestion. However, when the traditional traffic flow theoryresearches and describes the relationship, the effect of information elements is toosimplified, which hampers the accurate grasp of the relationship between drivingbehavior and traffic congestion, and is difficult to support the effective control method.In fact, all kinds of traffic physical objects are operated coordinately under thecomprehensive coordination effect of a variety of traffic information elements in theroad traffic system. The process has the typical characteristics of Cyber PhysicalSystems (CPS), especially in the traffic conditions of increasing road traffic density.The interaction between traffic physical system and traffic information system becomesmore and more important to the cognition of the relationship between driving behaviorand traffic congestion, and suppression of traffic congestion. Therefore, research on thetraffic congestion characteristics and the suppression method in CPS based on drivingbehavior has important theoretical and practical significance.Therefore, on the basis of the existing traffic flow theory and traffic congestionsuppression method, with consideration of the effect of information response delay andinformation anticipation factors on traffic flow (physical entity) in the process ofawareness, transmission, processing and decision making of traffic information. theauthors respectively study the relationship between traffic congestion and microanticipation driving behavior, macro response delay driving behavior from interactiveintegration between traffic information system and traffic physical system to obtain newmicro¯o driving behavior models for describing the microscopic formationmechanism and macroscopic propagation mechanism of traffic congestion respectively,and then explore micro¯o control strategies of road traffic congestion control inCPS.The main works of this dissertation are as follows:①From micro level, by considering the influence of the anticipation drivingbehavior on the evolution characteristics of traffic flow, a anticipation drivingcar-following model (ADCF model) is presented, then the evolution rule of headway wave of the ADCF model is obtained under the effect of anticipationdriving.In real traffic, a car’s speed adjustment not only relates to the headway of theconsidered car, but also depends on the anticipation headway of the considered car.Therefore, by considering the anticipation driving effect on traffic flow from tightconjoining between traffic information system and traffic physical system, a newmicroscopic anticipation driving car-following model (ADCF model) is presented onthe basis of OV model to study the relationship between anticipation driving behaviorand traffic congestion. The stability criterion is derived from linear stability theory todescribe the relationship between anticipation driving behavior and traffic congestion,and the mKdV equation near the critical point for describing headway wavetransmission rule is obtained through nonlinear analysis. The simulation results showthat the ADCF model can simulate practical traffic phenomena, such as stop-and-go,system critical phrase transition, etc., and its simulation results are more close topractical value than that of the OV model under the periodic boundary conditions andthe open boundary conditions. At the same time, the anticipation driving effect canenhance the stability performance of traffic flow, improve the threshold of densityabove which traffic flow will turn into congestion and reduce the effect scope ofcongestion, finally the optimal value range of anticipation parameter in the ADCFmodel is obtained under the open boundary conditions when viewing the minimumsmoothness and minimum fluctuation amplitude of speed as the evaluation index.②Based on ADCF anticipation driving model, by considering the influenceof the safe distance effect and comprehensive information of the nearest-neighborleading car on traffic flow in CPS, two improved micro congestion controlstrategies are presented respectively, and the validity of the two methods areproved by numerical simulation.Traffic congestion is the external performance of the unstable interior of roadtraffic system. In order to investigate the feedback control problem of traffic jam fromthe viewpoint of the system, the dissertation presents a new feedback control methodbased on ADCF anticipation driving model to suppress the traffic jam under openboundary condition, which considers the influence of the safe distance on traffic flow.Based on the feedback control theory, the stability criteria are provided as the speed ofthe preceding vehicle changes. Numerical simulations show that the congestion controlmethod considering a safe distance term can more enhance the anti-interference ability of traffic jam than that of the previous control method, effectively suppress or easetraffic congestion phenomenon.Based on the above results, by further considering the influence of optimal speeddifference information, speed difference information and the safe distance effect of theconsidered car, a new feedback control method considering comprehensive informationof the nearest-neighbor leading car is proposed. Based on the feedback control theory,the stability criteria are provided as the speed of the preceding vehicle changes, theoptimal value of the feedback gain is obtained under the open boundary conditionswhen viewing the minimum smoothness and minimum fluctuation amplitude of speedas the evaluation index. The theoretical analysis and numerical simulation results showthat the congestion suppression performance of the new method is better than that of themethod considering the safe distance effect. The simulation results are consistent withthe theoretical analysis.③From macro level, considering the influence of information response delaydisturbance on the macroscopic propagation mechanism of traffic congestion, anew response delay driving lattice hydrodynamic model(RDDLH model) ispresented, and the transmission mechanism of traffic congestion in unstableregions of new model is studied in detail.Based on Nagatani’s lattice hydrodynamic model, by considering the influence ofthe information response delay disturbance on the transmission mechanism of trafficcongestion from tight conjoining between traffic information system and traffic physicalsystem, an improved lattice hydrodynamic model (RDDLH model) is presented. Themodel’s critical stability condition is acquired through linear stability analysis. Thetransmission characteristics of density waves near the critical point are researched bynonlinear analysis method and mKdV equation is deduced. The simulation results showthat the response delay driving effect plays an important role in phase transition and asthe information response delay disturbance increases, the stable area becomes smaller,traffic jam becomes more serious, the density wave propagation velocity becomes fasterand the effect scope of congestion becomes bigger. This phenomenon is close to theactual traffic situation. Therefore, the RDDLH model more accurately reveals themacroscopic propagation mechanism of traffic jam than that of Nagatani’s model.④Based on RDDLH response delay driving model, by considering theanticipation effect of the nearest neighbor leading lattice and multiple precedinglattices on macroscopic traffic flow in CPS, two new macro congestion control strategies are presented respectively, and their congestion suppressionperformance are analyzed in detail.With considering of the anticipation effect of the nearest neighbor leading latticeon traffic flow from the perspective of tight conjoining between traffic informationsystem and traffic physical system, a new congestion control method consideringanticipation driving effect is developed. The results show that that anticipation effectcan effectively prevent traffic jam resulted from the information response delaydisturbance.Based on the above results, by further considering the influence of arbitrarynumber of preceding lattices on traffic flow, an improved congestion control methodconsidering multi-anticipation driving effect is proposed. The control model’s criticalstable condition is acquired through linear stability analysis. The mKdV equation fordescribing the control model’s density wave transmission rule in instability region isobtained by nonlinear analysis. Theoretical analysis and numerical simulation manifestthe suppression performance of the congestion control method consideringmulti-anticipation driving effect is better than that of the congestion control methodconsidering anticipation driving effect. Especially, the optimum steady state of thecontrol model considering multi-anticipation driving effect can be obtained by justconsidering the anticipation information of two preceding lattices.In summary, by deeply analyzing the influence of the anticipation information andinformation response delay disturbance on the physical state of traffic flow from tightconjoining between traffic information system and traffic physical system, buildingmicro¯o traffic flow models to precisely cognize the relationship between drivingbehavior and traffic jam, and essentially reveal the microscopic formation mechanismand macroscopic propagation mechanism of traffic congestion. Based on these models,micro¯o control strategies of suppressing road traffic congestion are explored inCPS. The achievements will contribute to well understand the mechanisms of roadtraffic jams as well as to propose a new architecture and methodology for congestioncontrol in CPS, and consequently support the development of Transportation CyberPhysical Systems. |
PDF Full Text Request