城市交通网络的控制与仿真研究-控制理论与控制工程专业论文.docxVIP

URBAN TRAFFIC NETWORK CONTROL: MODELING AND SIMULATION Abstract With the era of information, the step of the city life has become faster and faster, the conflict between the city transportation and economic development is gradually becoming critical. Traffic congestion, traffic security and environmental pollution have become the focus of social concern. As one of the primary methods to organize and control traffic flow, urban traffic control system plays an important role in solving the problems of transportation. But the traditional traffic control methods cannot solve the problem efficiently, and how to decrease the traffic press using modern science and technology is the emphasis of research in this field. For investigating urban traffic control strategies, a macro urban traffic model is at first improved and a distributed predictive control algorithm based on both traffic model and historical data designed on this basis. To solve the traffic congestion problem, we give the definition, description of traffic congestion and its solving method. The congestion removal control algorithm is applied to a single crossroad and urban traffic network. Summarily, the main research work of this dissertation lies on the aspects as follow: A control-oriented macroscopic discrete dynamic traffic flow model is discussed. A general urban traffic network topology is proposed by decomposing it into certain network elements. In order to verify the efficiency of above macroscopic model, we use least square method to compensate the error, which makes the accuracy of the macroscopic model more close to the real traffic situation. A distributed predictive control algorithm based on both traffic model and historical data is proposed and compared with decentralized predictive control algorithm based on this model. The distributed predictive control algorithm considers the interactions among crossroads and between neighbor green times, so it can get better control results. A new traffic