中文摘要：

针对中国大城市的工作生活模式与公交网络的特性，以成都市居民公交出行为研究对象，提出了符合乘客路径选择行为的广义公交路径。考虑了公交出行的路段阻抗和站点阻抗，建立了公交路径阻抗函数，提出了有效路径的确定方法。基于改进的Logit模型，建立了一种换乘行为影响下的路径选择模型。以成都市部分公交网络为例，应用提出的配流算法进行实例验证。分析结果表明：当选定的4条公交线路高峰时段的最小发车间隔分别为4、4、3、3min，非高峰时段的最大发车间隔均为10rain时，对应的最小换乘步行时间和最大换乘步行时间分别为0、6rain；当最大路径阻抗和最小路径阻抗分别为71．5、51．5min时，对应的乘车时间分别为60、48min，路径选择比例分别为5．53％、41．98％；当最大路径阻抗和最小路径阻抗分别为52．5、48．5min时，对应的乘车时间分别为42、39min，路径选择比例分别为13．40％、62．07％；考虑换乘行为时，配流结果与实际值的最大相对误差、最小相对误差和平均相对误差分别为16．46％、11．09％、14．42％，不考虑换乘行为时，最大相对误差、最小相对误差和平均相对误差分别为34．37％、11．38％、23．15％。考虑换乘行为的配流结果更贴近实际情况。

英文摘要：

Aiming at the work and life modes and the particularities of public traffic networks of big cities in China, the public traffic trip behaviors of residents in Chengdu City were taken as study object, and the generalized public traffic route which could accord with the route choice behavior of passenger was put out. By considering the section impedance and station impedance during the public traffic trip, the impedance function of public traffic route was set up, and the determination method of effective path was proposed. Based on the improved logit model, a route choice model under the influence of transfer behavior was established. The part public traffic network of Chengdu City was taken as an example, the example verification was carried out by using proposed assignment algorithm. Analysis result indicates that when the minimum departure intervals of four designate public traffic lines during peak period are 4, 4, 3, 3 min respectively and all the maximum departure intervals during non-peak period are 10 min, the minimum and maximum transfer walking times are 0, 6 min respectively. When the maximum impedance and minimum impedance are 71.5, 51.5 min respectively, the riding times are 60, 48 min respectively, the route choice ratios are 5.53%, 41.98% respectively. When the maximum impedance and minimum impedance are 52.5, 48.5 min respectively, the riding times are 42, 39 min respectively, the route choice ratios are 13.40%, 62.07% respectively. When the transfer behavior is considered, the maximum relative error, the minimum relative error and the average relative error of assignment results and actual values are 16.46%, 11.09%, 14.42% respectively. When the transfer behavior is not considered, the maximum relative error, the minimum relative error and the average relative error are 34.37%, 11.38%, 23.15% respectively. The public traffic assignment result considering transfer behavior is in accord with actual situation. 2 tabs, 10 figs, 24 refs.