用“多点摄像法”对长春市的2个路段交通流进行了实地调查。通过数据处理，给出了车队离散过程中的多点流量图式以及多段的车速分布。用实际观测数据所得的流量图式与TRRL的流量图式进行对比，分析异同。还对车辆行驶速度及行驶时间分布进行了拟合，分析了各离散模型的适用范围。最后对车队离散进行计算机仿真，得出了满意的结论。

介绍了交通流理论的研究现状，指出目前交通流理论研究的重点和难点是网络交通流的动态特性以及路网交通流点、线、面之间的动态联系。通过对路网交通流特性及电路网络特性的对比分析，论述了用电路网络模拟道路网络来研究交通流的动态特性和联系特性的可能性。

Based on the relative relationships between the traffic volumes of intersections, we do main-component analysis to the intersections with detectors and create mainfactors. Then we utilize these main-factors to predict the traffic volume of nondetector intersections and the results are examined by the traffic volume data of the road network of Changchun City. The problem of the traffic volume prediction of non-detector intersections is studied in this paper, the obtain result can be widely used in urban traffic flow guidance in intersections where has no enough detectors equipped. Some theoretical basis for macroscopic management is also provided by this method.

In terms of the similarity between a road network and a circuit network, this paper demonstrates that Kirchhoff's current and voltage laws observed in the circuit can be employed in the traffic assignment of the road network. Inspired by the more consummate circuit theory, we put forward a new idea and method of traffic assignment. A set of models is formularized by means of multiobjective nonlinear planning theory, and the solutions are found using MATLAB. An example is provided to prove the feasibility of the new traffic assignment model.

介绍了一种新的计算饱和流率和相位损失时间的累计曲线法。与韦伯斯特方法相比，该方法可以计算前损失时间和后损失时间，弥补了韦氏方法的不足。实际应用结果表明，在无干扰或干扰很小的情况下，一条车道的饱和流率在1600～1900veh/h之间，总损失时间在1～2s之间，有的相位损失时间可以忽略不计；饱和流率的计算值与通常假设的1700～1800veh/h基本相符，但实际测得的损失时间却低于HCM推荐的3～5s, 也低于黄灯时间。