Study On The Impacts And Dispersion Of Traffic Incidents On Urban Roads

The current frequency of traffic incidents on urban roads remains high, which has been one of the major reasons for causing traffic congestion and reducing operational efficiency of the road network. By exploring the diffusion mechanism, dispersal and control strategy of traffic incidents on urban roads, the traffic congestion can be prevented and controlled effectively. By constructing traffic flow model of urban network under the traffic incident, the operation rules for traffic flow are analyzed, the formation and diffusion mechanism for the impact of traffic incident under different road network is revealed, and the strategies for relieving traffic congestion under traffic incident on urban roads are explored.First, based on the cell transmission model and the detector data under the traffic incident, the traffic flow model of road network under the influence of traffic incidents (ECTM) is builded by considering traffic density as the state variables. And in this model, the way of cellular division is adjusted for different locations of traffic incident, by which the analysis for the incident sections and normal sections of road is performed differently, in addition to it, the sending and accepting functions are redefined by considering the impact of traffic incident.Based on ECTM, the concept of the impact length for the incident is presented, which is for the quantitative characterization of the influential region of the traffic incident. And combined with travel time, average congestion delay, average travel speed and impact length, an impact evaluation system for traffic incident is builded. By considering the impact of intersection channelization, signal timing and road opening on traffic flow, the range forecast model of traffic incident impact under ordinary urban road (ECTM-R) is constructed, in which the simple handling of nodes in the CTM is avoided, and a true reflection for the traffic transmission of cells under different connection is presented. The model is verified by the measured data from the main city zone of Nanjing, and it is showed that the trends of the measured data and the simulation data are basically the same. To further verify the accuracy of ECTM-R, the impact assessment indicators and error of the model is compared with existing models (SCTM model) by simulation analysis, and it is showed that the accuracy of ECTM-R model is higher.In the link model, the typical phenomenon of traffic flow on urban expressway is portrayed by taking the impact of metastable state on traffic flow into account; in the node model, the impact of the entrance ramp on traffic propagating is considered. Based on these, the range forecast model of traffic incident impact under the urban expressway (ECTM-F) is constructed by combining with ECTM model. Through the instance analysis of local expressway network in Beijing and the comparative analysis with SCTM model, the results show that the trends of the simulation data obtained from ECTM-F are basiclly the same with the measured data and the accuracy of this model is higher than SCTM Model. Respectively based on ECTM-R and ECTM-F model, the traffic state of urban expressway and ordinary urban roads under traffic incident is studied at peak periods and off-peak periods, and the differences between the incident impacts under the two different network conditions are described. The results show that, when traffic incident occurs in the expreeway at off-peak periods, the congestion delays, the growth rate and the decline rate of the influential region of traffic incident are all higher than that of ordinary urban roads; when traffic incident occurs in urban freeway during peak periods, the growth rate of impact evaluation index is still higher than the ordinary urban roads, but the decline rate is lower than the ordinary urban roads.The predicted influential region of traffic incident on ordinary urban roads can be divided into three areas based on ECTM-R model:disposal zone, control zone and warning zone. According to different regional characteristics, different dispersion purposes and dispersion strategies are proposed. On this basis, considering the control zone as the research object, the traffic dispersion model is established by ant colony algorithm and orthogonal design, in which the homodisperse and homogeneous of orthogonal design can effectively make up for the deficiencies of the ant colony algorithm, which are long search time and falling into local optimum solution easily. Combined with the measured data in the selected road network, traffic assignment scheme can be quickly generated by this model. And compared with the existing model, the results show that, the traffic assignment scheme acquired by the model based on orthogonal design and ant colony algorithm is more balanced and reasonable.By considering the basic characteristics of traffic incidents on urban expressway, the overall objective of traffic dispersion for incidents is proposed. Based on this objective, combined with ECTM-F model, the division of influential region for traffic incidents is determined, and the dispersion of different disposal area based on VMS are presented. The strategies are applied to a selected road network, and the results show that the strategies can suppress the impact of the traffic incident more effectively by compared with the existing method.