采用一维和二维元胞自动机（CA）模型对于城市与高速公路中的交通流进行微观高速模拟，通过颗粒流体动力学分析、统计力学处理、精确平均场方程研究和非线性科学方法讨论，研究交通运输复杂系统的自组织临界性和相变行为。分析各种真实交通条件的非线性效应，发展更有应用价值的交通流模型，为建立先进的交通运输信息管理系统提供理论依据。

We propose new types of traffic flow cellular automaton models for the car that can only be accelerated gradually and for the car that can be delayed stochastically only when following the trail of the front car. Using the spacing oriented mean field analytical approach, we present the exact analytical fundamental diagrams for phase transition of the new models. We propose two way choice decision traffic flow model for the intellectual vehicles, study the mechanism affecting total traffic flux and stability of the whole two-way system. We propose traffic flow models with anticipation of movement of the car ahead, presenting the characteristics more consistent with the observation on real traffic flow. We study the on-ramp system using the CA traffic flow model, considering not only the influence of the on-ramp flow on the main road but also the opposite influence, hence obtain the basic understanding for the formation of synchronizing traffic flow. Besides, we study chaos controlling and chaos synchronization, forest fire model and self-organization criticality, the financial market models based on agent interaction and the phase transition behaviors of general complex adaptive systems. All the 20 papers published in international journals with English and 24 papers published in national journals with Chinese and three invited parallel reports in the international conferences were supported by this NSFC foundation.