Urban Traffic Network Design Involving Signal Control Optimization

With the steady development of the social economy and the rapid propulsion of urbanization in recent years, the contradiction of urban traffic supply and demand turns worse, the traffic problem has become one of the main factors that hinder the development of cities. For this reason, it is a very pressing issue for us to have a systematic study of urban traffic network design problem so that it can guides the actual traffic network reconstruction projects and promote sustainable urban transportation development.Considering the traditional traffic network design research have failed to analyzing the influence of traffic signal control to urban road network traffic flow, this thesis synthesizes dynamic signal optimization control and traffic network lanes allocation, which is more accord with current situation of the urban traffic network.The main results of the thesis are as follows:1、In view of the Daganzo’s original cell transmission model (CTM) is obvious insufficient in the treatment of such case just like multi-cellular merge, diverge and multiple-twin intercellular traffic streamlines cross, this thesis puts forward a new traffic flow simulation method based on the accumulative staged cell transmission flow(ASCF-TS) which not only meet the intracellular traffic flow FIFO criterion and intercellular transmission capacity constraints, but also satisfy the transmission capacity constraints of traffic streamlines intersection, meanwhile it can deal with the priority problem among the crossed traffic streamlines availably, enhancing the CTM model applicability.2、For the implementation of traffic signal dynamic optimal control, a cell-based variational inequality formulation of the dynamic traffic assignment is adopted to indicate the urban road network users’path selecting, and an alternating direction method is devised to solving this problem according to the path travel time monotone mapping path traffic flow but not differentiable.3、After analyzing two kinds of optimization variables including lane numbers and signal parameters (the cycle length, phase’s green time),as well as two kinds of objective functions including the minimum total traffic delays and the maximum network reserve capacity, a multilevel traffic network design model is constructed. By solving it, traffic network lanes allocation optimal scheme which has the universal applicability for all or some kinds of representative network traffic patters and signal control parameters optimal scheme which enforces dynamic adjustment with the dynamic change of traffic.4、A nested genetic algorithm is devised for solving this non-convexity multilevel network design problem, it’s fitness functions, Individual chromosomes, decoding method and selection operator, cross operator and mutation operator were designed concretely. The final numerical example’s results in chapter6verify the rationality of the model and the validity of the algorithm proposed in this thesis.