Primary Study On Traffic Flow Chaos

Research in chaos stands at the front of nonlinear science. Chaos is a basic subject; chaos theory and chaos method can be used as a tool for the other subjects. The definition, the basic features and verification of chaos are detailed presentated in chapter one. The uncertainty of traffic system is in close connection with chaos. In chapter two, several factors causing the uncertainty of traffic flow are discussed. Besides, the present state and the problems in chaos research are reviewed based on the analysis of the references. In chapter three, a new approach for evaluating of reasonability for the traffic flow models has been presented. The time serials of traffic flow are generated by several most common used car-following models. The appearing of chaos state can be taken as a criterion of the reasonability of the traffic model. This approach presents a new idea for evaluating for traffic flow models.The main topic of chapter four is to find the nature of chaos in traffic flow. The research deal with different aspects of traffic flow get from different way. 1.Analysising the traffic flow data at Weijin road and Honqi road in Tianjin with chaos criterion. 2.Analysising the time series generated by traffic software with chaos criterion. 3.Analysising the time series generated by traffic flow models with chaos criterion. Farther experiments compared the chaos states of these three kinds of typical traffic flows by their correlative dimension and Lyapunov exponents. Results show that the three kind of traffic flow are similar in chaos nature. The tiny difference between the results originates from the simplifying of the actual circumstance in simulation software and traffic model. Certainly, academic researches are aimed at practical application in engineering. Chapter five presents several approaches for chaos controlling in traffic flow. How to keep the stabile and orderly state of traffic flow is an engineering problem. One of the effective ways for well traffic control is to bate or weaken chaos state. Up to now, there are two kinds of appoach in chaos control: one is feedback control , and the other is no feedback control. Car-following models are taken as subject in the research. As to the first method, by adjusting correlative model variables with feedback method, chaos can be eliminated. The other two no feedback methods: control through parameter and control through disturbance, are presented in the paper too. By comparison, method of control with feedback is prone to be realized in engineer. However, the last two methods is a valuable attempt. With developing of corresponding theory and technology, chaos theory will see a more flouring future in not long time.