中文摘要：

为了提供一种多原则指导下的交通流建模及求解方法,针对非两类极端条件下（用户最优、系统最优）的交通流分配问题展开研究,着重解决追求系统总体最优而忽视个体出行者承受多余开销导致的服从意愿较差的问题。引入距离和时间2种因素重新定义了车辆综合开销,从增强诱导信息影响效果的角度,将个人因服从诱导信息所新增的开销引入动态分配模型的约束条件中,以一种启发式的免回溯求解算法降低计算复杂度,并分析了近似比。研究结果表明：模型在以系统最优为主要目标的前提下可帮助出行者以较低的绕行开销避开拥堵,提高用户对诱导信息的遵从程度;在5类指标方面的表现显著优于DUO及最短路径模型;近似算法以接近线性的计算复杂度达到了DSO系统最优模型90%以上的效果。

英文摘要：

For providing a method of traffic flow modeling and its solution under multi principles,a research on traffic assignment problems under two non-extreme conditions（not user optimum and system optimum）was carried out.The key point was to solve the problem of ignoring individual traveler＇s poor compliance due to compulsory extra cost in the pursuit of overall system optimization.The introduction of 2factors including distance and time redefined the vehicle integrated cost.From the perspective of enhancing guidance information effect,newly increased cost of individual resulting from obeying the guidance was introduced into the constraint condition of the dynamic assignment model.And a free heuristic non-backtracking algorithm was put forward to reduce the computational complexity,and the approximate ratio was analyzed.The results show that the model can still help travelers avoid congestion in low circumambulate cost and enhance user＇s compliance to traffic guidance information under the premise of system optimization as the main object.The model in this paper has a significantly better performance than DUO and the shortest path model in 5 indexes. The computational complexity of approximation algorithm is nearly linear but it reaches more than 90% effect of the DSO system.