相比基于均衡原理所建立的网络均衡配流模型，日变交通配流模型能够描述网络系统的非均衡状态，因而具有更一般的适用性． 针对基于期望效用理论的出行者路径选择模型的不足，采用累积前景理论获得了出行者具有参考点依赖特征的路径累积前景值．假设出行者是最大路径累积前景的追随者，利用不同路径累积前景的差异构建了网络系统路径流量交换模型，并据此建立了一个日变动态配流模型． 通过一个算例网络对配流模型进行了验证，数值结果表明所提出的模型能够反映动态决策环境下出行者考虑随机风险的有限理性决策行为．

As compared with the network equilibrium assignment model based on the equilibrium principle,theday-to-day traffic assignment model is more feasible due to its capability to describe the disequilibrium status of anetwork system.In order to overcome the defects of the expected utility theory-based route choice model of trave-lers,the cumulative prospect theory is used to obtain the cumulative prospect value of the traveler route with thecharacteristic of reference point dependency.Assuming that travelers are the followers of the maximum cumulativeprospect of routes,a route switching model for network systems is constructed according to the discrepancy betweendifferent route cumulative prospects,and then a day-to-day dynamic assignment model is proposed.Finally,theproposed assignment model is verified by a numerical test on a network.The results show that the proposed assign-ment model can describe travelers' boundedly rational behavior considering stochastic risks in dynamic decision-making contexts.