Green Time Model Free Adaptive Control Of Isolated Intersection

With the development of the economy and urbanization, many cities in china have suffered traffic congestion, because of the rapid development of urban road infrastructure and the fast increasing traffic demand. Congestion pricing is an effective measure to solve congestion in urban road, which is widely used in many controies all over the world. How to choose a reasonable fee is a key problem in congestion pricing. At present, most congestion pricing models are static model with double-layer and dynamic model based on dynamic assignment and non-double-layer. Static pricing strategy is fixed and not real-time variable, which is not an optimal pricing strategy to alleviate the traffic congestion. Although dynamic pricing strategy of dynamic model based on dynamic assignment and non-double-layer can be real-time variable in traffic flow, the model is non-double-layer, only considering the dynamic user equilibrium and ignoring the upper management system.In this paper, a double-layer model of dynamics of charge in congestion pricing is proposed, which takes into account not only dynamic assignment of flow, but also user equilibrium and system optimal. Firstly, double-layer model of dynamics of charge in congestion pricing is proposed, basing on mathematical programming with double-layer theory and the existing fixed static demand charges model and dynamic allocation of flow theory. The low of the model established meets the first principle to the requirements of Wardrop - stochastic dynamic user equilibrium. Considering the current time of arrival based on traffic flow, users randomly select the shortest path to reach the destination at the decision-making node. Top model is to minimize the cost of the total time to achieve the objective of reducing the congestion in the study period. Secondly, the MSA algorithm is developed to solve the low model to meet stochastic dynamic user equilibrium, the flow solution will be incorporated into the top model, and use the algorithm of the step acceleration and penalty function method to gain the best pricing strategy. Finally, some experimental studies have proved that the proposed dynamic pricing strategy is effective for alleviating the problem of urban road congestion.