Study On Coordination Control In Urban Traffic Signal Networks

The thesis makes a comprehensive and deep study on coordination control in urban traffic signal networks, which was based on the west region of Zhongguancun of Beijing. It is through adjusting the parameters of signal control and coordinately controlling of the urban traffic flow by real-timely monitoring the traffic flow and real-timely optimizing the traffic control model to reduce the stop times and delay, and aiming at improving the performances of urban traffic networks.The paper including:1. A kind of flexible two-stage decomposition fuzzy controller for urban traffic coordination control is put forward. There are some differences in traffic capacity and function of each road line in the urban traffic network, for that reason the urban traffic network is separated into main trunks and branch trunks. In this paper a kind of flexible two-stage decomposition fuzzy controller for urban traffic coordination control is put forward. The fuzzy controllers of control level are designed for all intersections in the urban traffic network and the fuzzy controllers of coordination level are only designed for the conjoint intersection in the main trunks. The controllers of coordination level can coordinate the control signal of the conjoint intersection in the main trunks. This kind of control structure decreases the number of fuzzy rules greatly, improves the operation speed and the traffic capacity of the main trunks, and consequently improves the traffic environment of the whole road network.2. As the traffic system has nonlinear, fuzziness and uncertainty, is a distributed system. A kind of urban traffic control system based on Multi-Agent is designed in this paper. The control system adopts the distributed control structure and distributed compute mode which is to strengthen the real-time property and robustness of urban traffic control system.3. A kind of urban traffic coordination control arithmetic based on distributed Q-learning is designed for the distributed control structure of Multi-Agent. And designed a kind of reward function and distributed weight function, which are fit to the fields of traffic control. The urban traffic coordination control arithmetic based on distributed Q-learning doesn't need an upper supervision level which is widely used in traditional urban traffic control. It is a kind of potentially instructive urban traffic coordination control mode which doesn't need traffic flow model and the...