The Research Of Urban Traffic Management Strategy And Simulation Based On Game Theory

The control of traffic can be mainly divided into two parts, the control ofvehicles and the control of intersection signals. The traffic manager can takeadvantage of them to realize the safety and smooth of roads effectively. But in thereal situation of traffic management, traffic management is often unable to takeeverything into consideration because of the restrictive effects among the vehiclesand among the intersections. This paper finds out a balance among them by utilizinggame theory and analyzing the effects among the traffic objects. A simulation isimplemented to test and verify the control effects. The main contents of this paper areas following.(1) This paper introduced a nonlinear dimensionality reduction method based onmanifold learning, namely the Local Tangent Space Alignment, into the field offeature extraction in urban traffic networks. By constructing an approximation for thetangent space at each data point in high dimensionality, this method is capable oflearning the local geometry of the manifold. Those tangent spaces are then aligned togive the global coordinates of the data points in a lower-dimensional space withrespect to the underlying manifold. Experiment shows that compared with otherdimensionality reduction methods such as Principal Component Analysis and LocallyLinear Embedding, this manifold method has a more impressing performance inextracting as well as visualizing the temporal-spatial features of the traffic network.(2) To analyze the interplay of the crossings in the traffic by studying the signalcontrol. The multi-intersection signal control is always the difficult part in the trafficcontrol. In this paper, each intersection is deemed as an independent autonomousindividual. By studying the restraint among the intersection signals, the balance pointwill be found by using game theory. Firstly, a road network is built by using thesimulation software VISSIM, and then traffic status information is extracted and processed. Secondly, based on the past data which is used to predict the traffic flowinformation, we find the optimized point after a certain time of traffic control byusing the game tree, and trace back to the root node to obtain the optimal strategy ofeach intersection on current conditions. Evaluation function in the game tree will betaken into account to evaluate the influence of other intersections to the targetintersection to get the best game equilibrium solution.This paper uses Game theory to find the equilibrium among all the trafficparticipants, the control strategy based on game theory can adapt all the traffic state,it’s helpful to improve the traffic safety and efficiency.