一种拥挤数量调节的非均衡网络交通流动态演化模型（英文）

基于经济学非瓦尔拉斯均衡理论, 采用经济学中市场摸索过程模拟出行者路径选择行为; 假设城市出行者在路径决策过程中, 考虑路径出行时间和关键路径拥挤程度的共同影响, 以价格拥挤混合均衡交通流模式为基础, 建立了一种价格-拥挤混合调节的非均衡网络交通流动态演化模型, 并验证了模型稳定状态与均衡的等价性; 基于简单的测试网络和中型路网, 对演化模型进行了模拟, 描述了非均衡网络交通流的演化过程与非均衡状态下交通网络的整体表现。研究结果表明: 时间价格调节模型的演化结果符合经典的Wardrop第一原理, 拥挤数量调节的结果使得OD间各路径上关键路段的拥挤程度一致, 价格-拥挤混合调节的结果会使路径流在走行费用较小和拥挤程度较低的路径上相互进行调整, 其动态演化过程波动性要大于单一调节的情况; 在测试路网中, 考虑采用拥挤程度对路径进行选择的行为, 使得整个路网拥挤均匀程度整体提高62%, 但路段饱和度均值却从0.60增大到了0.64, 表明路网整体上变得拥挤; 若考虑两者的共同调节, 最拥堵路段饱和度从0.936下降到0.787, 均匀程度整体提高46%, 且路段饱和度均值降低, 路径行程时间变小, 拥堵得到改善; 中型路网的测试结果也表明这种混合均衡模式能灵活、客观地描述路网交通流动态演化过程, 获得较为合理的路网系统的稳态流量。 

A dynamic evolution model of disequilibrium network traffic flow with quantity regulation of congestion

Supposing that the travel cost on the paths and the congestion degree on the key links were considered by the urban travelers, aprice-congestion mixed dynamic evolution model was established based on analyzing the equilibrium flow model.The model was based on the economics theory of non-Walrasian equilibrium method and by simulating the traveler's route choice behavior following the economical concept of market exploration process, the equivalency of model stability and equalization was verified.The evolution model was simulated by using a simple test network and a medium size network, the evolution process of disequilibrium network traffic flow and the performance of traffic network under the disequilibrium situation were described.Analysis result indicates that the evolution model of time price regulation accords with the classical Wardrop's first principle; the result of quantity regulation of congestion allows the degree of congestion on the key links of each path between OD to be the same; the result of pricecongestion mixed regulation allows the path flow to be adjusted between the paths of lower cost and the ones of less congestion, the undulation of dynamic evolution of which is greater than that of the single regulation.In the test network, because the model only considers the choice behavior of congestion degree upon path, the congestion degree of whole traffic network is more uniform, and compared with the single price regulation model, the overall uniformity coefficient improves by 62%.However, the mean link saturation improves from 0.60 to 0.64, which indicates that the traffic network becomes congested overall.By considering the joint regulation of these two factors, the saturation of most congested link decreases from 0.936 to 0.787.The overall uniformity coefficient improves by 46%.The mean saturation of links, path travel time and congestion decrease.The test result of the medium size network also shows that such mixed equilibrium model can describe the dynamic evolution process of traffic flow on traffic network flexibly and objectively, and achieve steady state flow of traffic network system, which canexplain the traffic travel behavior better.8 tabs, 9 figs, 32 refs.