中文摘要：

为研究多类型出行者在实际路网上的路径选择行为，提出了动态广义出行费用概念，假设路网随机降级交通需求为弹性，并且出行者以感知的最小广义费用为出行依据，在此基础上构造了多类型随机用户平衡模型，设计了相继平均算法求解．在一个测试路网上进行了算例分析，结果表明：熟悉路况的出行者容易选择到费用较小的路径；路况熟悉程度、权重系数，以及规定阈值均对平衡分配结果有很大影响，且结果与实际经验相符合，验证了模型的实用性．

英文摘要：

To study route choice behavior of multiple class travelers on actual road network, the conception of dynamic generalized travel cost was proposed. There are some assumptions that road network are degradable, traffic demand are elastic, and travelers choose their optimal routes while their generalized travel costs are minimized. On this basis, a stochastic user equilibrium model with multiple classes was established, and a method of successive average （MSA） was designed to solve the model. The analyses of numerical example were conducted on a test road network, and the results show that travelers who familiarity with road conditions are easy to choose the less cost route, and degree of familiarity with road condition, weight coefficient and given threshold have great influences on traffic assignment. The results are consistent with actual experiences, so the practicability of the model is verified.