Traffic Assignment Model And Algorithm With Traffic Management And Control For Congested Networks

With the intention to optimize the traffic assignment model to better reflect the facts in road network, user equilibrium model is built with turning delays and link capacity restraint, considering the influence of intersection delays and link capacity on traffic assignment. The algorithm of this model can get path information, providing references for the optimization of traffic management measures.This thesis firstly introduces the conditions of traffic management and control, which mainly include management conditions of road section and intersection; Secondly, states the features and scope of application of traffic management and control. Joint with the characteristics of transportation network flow theory, these conditions are processed mathematically. In the traffic management conditions, the function values of different road sections' free time are set to control the road section management method of one-way traffic or forbidding. The turning delays conditions are involved in the intersection management mode to describe the traffic management measures.Based on the path, user equilibrium model is built with turning delays. The Projection Gradient algorithm is more efficient than the classical Frank-Wolfe algorithm. The example analysis shows that this user equilibrium model can better screening influence of intersection delays on traffic assignment. Besides, The Projection Gradient algorithm can get path solution, and can confirm the traffic origin and destination of a certain road section. Through the constrain of a certain diversion, the traffic volume transfer situation can be clearly known, providing traffic management and control with scientific decision aids.The traffic distribution model is built with turning delays and link capacity constraints. In the aspect of solving algorithms, internal recycle uses Projection Gradient algorithm, external recycle uses Augmented Lagrangian Multiplier method. The feasibility of the algorithms is tested by examples which show that after adding capacity constraints, road traffic flows are all within the scope of the link capacity, which is an ideal state of traffic flow distribution. This method provides scientific directions to the analysis of traffic congestion state and traffic improvement.This thesis also includes the optimization of traffic management measures. With the example of Sioux Falls network, in the road networks which need the consideration of turning delays, use path and link information to adopt corresponding traffic management methods, and alter the traffic flow distribution states of the whole road network to make it more similar to those with consideration of intersection turning delays and link capacity constraints. Examples illustrates that after the use of proper traffic management, the number of oversaturated link can be reduces effectively, approaching to the distribution states with capacity constraint, thus can availably ease traffic congestions.