A Multimode-Based Dynamics Modeling Method For Urban Road Traffic Networks

With the rapid city development, various kinds problems emerging in urban road traffic systems, such as increasing jams etc., decrease the daily operation efficiency of urban road traffic systems. To solve those problems provides new issues for the research of complex system theory and urban road traffic networks.The character analysis of the whole behavior of specific urban road traffic systems is one of the important and hot topics, which can be described as the analysis of the formation and evolving processes of emergent behavior in complex systems, and the analysis of topological dynamics in urban road traffic networks. It helps us to find the essential characteristics and the development trend of the problems in urban road traffic systems from the perspectives of different abstract level.Complex system theory is and still will be for a long time an important issue in many specific research domains. Complex system theory aims at explaining the mechanisms and laws on which higher-level emergent phenomenon is generated by lower-level components and interactions among them. The critical point of the current research of complex systems is to understand complex systems and to formally define related concepts. Furthermore, emergence is the most intrinsic characteristics of complex systems. And understanding emergence is the main task in understanding complex systems.In this dissertation, based on comprehensively understanding the basic character of complex systems, we define emergence and related concepts that explain the forming process of emergence. According to our thinking about the forming and evolving process, the multimode-based complex system dynamics modeling method is presented. The method determines the correlation coefficient of each system mode that is denoted by a corresponding emergence attractor at a time phase, by simulating system behavior evolving processes under each system mode's affection only. The multimode-based complex system dynamics modeling method provides a new promising approach to study the dynamic characteristics of the evolving process of emergent behavior in complex systems, but also a modeling and analyzing tool for specific systems.As we known, large-scale urban road traffic systems are kind typical complex systems, including all research topics in complex system domain, and provide a platform for related study. The network-like models are always adopted in urban road traffic system domain; as well, complex network theory is often used in urban road traffic systems. As for network-like models, we introduce traffic level-of-service (LOS) described traffic states on roads, into urban road traffic networks to construct the variable-structure dynamic network model and the extended variable-structure dynamic network model. Based on the traffic data from the Beijing road traffic network, the structural dynamic characteristics of the extended variable-structure dynamic network model is validated; and the experiment shows that the model has three interindependent system modes:the static network mode, the stochastic network mode and the scale-free network mode. The three system modes define the corresponding evolving mechanisms of system behavior respectively, also called sub-dynamics.The modes'functioning patterns are concluded into two classes:the single pattern and the union pattern. Therefore, we construct the sub-dynamics models of the three system modes, according to which the evolving processes under three modes can be achieved respectively. In the correlation coefficient equation, the average node degree and the scaling exponent of node degree distribution are considered.In this dissertation, we select five time phases for the urban road traffic network dynamics modeling and analyzing experiments based on the real traffic data from the Beijing road traffic network. The experimental results show that the Beijing road traffic network has the multimode feature in a 24h time cycle; the major mode can characterize the Beijing road traffic network at each time phase; the multimode-based dynamics modeling method can quantitatively analyze the behavior of urban road traffic network, and point out the major mode and the mode functioning pattern.In conclusion, the multimode-based complex system dynamics modeling and analyzing method provides a new tool to both dynamics modeling of urban road traffic networks and evolving mechanism analysis of emergent behavior in complex systems.