基于综合交通网络的干线公路客流预测方法

提出了一种融合多种运输方式的干线公路客流预测方法；通过引入基于“人次”的标准客运单元和“点-线”的枢纽节点转化方法，将公路、铁路、航空以及水运等不同运输方式子网络进行融合，构建了可体现不同运输方式之间换乘关系的综合交通网络模型；考虑出行经济费用、出行时间、最大出行恢复时间、舒适度等因素，构建了综合交通网络下不同运输方式的阻抗模型；利用额定载客数和单位时间发车次数等参数，实现了综合交通网络下不同运输方式路段最大容量的标定；基于标准客运单元和综合交通网络模型提出了考虑综合交通阻抗的客流分布预测模型，实现了考虑其他运输方式影响的干线公路客流预测，并以黑龙江省哈大绥齐地区为例进行方法验证。研究结果表明：与2019年的实际观测值相比，在无伴行线路时基于综合交通网络的干线公路客流预测方法预测结果平均误差为5.47%，略低于传统四阶段法的6.14%，但在有伴行线路时该方法平均误差为4.58%，远小于传统四阶段法的11.89%；相比传统四阶段法，该方法能够更好地反映综合交通网络结构变化后转移客流对干线公路客流量的影响；相比新增水运线路，新增高速铁路或普通铁路伴行线路对干线公路客流影响更大，更能促使公路客流向铁路进行转移。

Trunk highway passenger flow forecasting method based on comprehensive transportation network

A trunk highway passenger flow forecasting method integrating multiple transport modes was proposed. By introducing the standard passenger transport unit based on the man-time and the hub nodes conversion method from point to line, the sub-networks of different transport modes, such as highways, railways, airlines and waterways, were integrated, and the comprehensive transportation network model which can reflect the transfer relationship among different transport modes was built. By considering the travel economic cost, travel time, maximum travel recovery time, comfort level and other factors, the impedance functions of different transport modes in the comprehensive transportation network were constructed. The maximum capacities of different transport modes in the comprehensive transportation network were calibrated by using the rated passenger number and the number of departures per unit time. Based on the standard passenger transport unit and comprehensive transportation network model, the passenger flow distribution forecasting model considering the impedance of comprehensive transportation was proposed. On this basis, the passenger flow forecasting model considering the influence of different transport modes was realized. Taking Harbin, Daqing, Suihua and Qiqihar area in Heilongjiang Province as an example, the method was verified. Analysis results show that compared with the actual observation value in 2019, the average error of forecasting results of the passenger flow forecasting method based on the comprehensive transportation network is 5.47%, slightly lower than the 6.14% of the traditional four-stage method when there are no accompanying lines around the characteristic roads. However, the average error of forecasting results of the proposed method is 4.58% when the accompanying lines are around the characteristic roads, far less than the error value 11.89% of the traditional four-stage method. Compared with the traditional four-stage method, the proposed method can better reflect the influence of the transfer passenger on the traffic volume of the trunk highway after the structural change of comprehensive transportation network. Compared with adding waterways, adding high-speed or conventional railways accompanying lines have more obvious impact on the passenger flow of trunk highways, and can promote the transfer of passenger flow from highways to railways. 4 tabs, 12 figs, 32 refs.