曹梦菲

城市空间紧张、道路交通拥堵、环境污染等问题日益突显，作为一种绿色、经济、集约的机动化交通方式，公共交通已成为优化土地利用模式、缓解交通拥堵和实现城市交通可持续发展的关键路径。然而，受到私人机动化出行的冲击，以及公共交通自身服务水平不高、竞争力不足等问题的影响，公共交通发展遭遇到严重的瓶颈，面对居民出行需求日益多元化，公交出行从“线路出行”不断向“网络出行”方向转变的现实环境，公交线路优化理念需要实现从“被动适应”向“主动引导”的转变，因此，把握出行者日常活动与公交需求的特征，探索公交站点的引流机制及分级特征，解析公交网络的拓扑结构并进行网络重构优化成为创新公交服务模式的关键。鉴于轨道交通对于公交网络的锚固作用，本文进行公交网络优化时主要从常规公交进行考虑。

本研究首先从时空两个维度分析了居民活动分布特征、产生机理和常规公交站点层的公交客流特征。采用居民出行调查数据对出行特征进行分析；进而综合利用手机信令数据和居民出行活动调查数据，对居民日常活动的时空分布特征进行探究；并以活动OD为切入点，分析活动流向的特征变化；结合POI数据，从用地角度解析活动空间分布特征的产生机理；利用常规公交IC卡数据和线路站点地理信息数据，对研究区域内工作日和非工作日公交客流的时间和空间分布特征进行分析，挖掘站点客流的空间特征变化情况。

其次，利用复杂网络理论，通过站点重要度评价进行关键站点识别与站点区域等级划分。构建现有常规公交网络拓扑模型，对网络的结构特征进行多角度分析，并考虑空间位置和活动需求，建立多维度站点重要度评价指标体系，采用加权TOPSIS（Technique for Order Performance by Similarity to an Ideal Solution）算法进行网络关键站点的识别，进而根据站点重要度得分，利用K-means聚类算法对站点重要度进行等级划分。基于站点重要度和空间位置特征，利用Geohash空间搜索算法，对不同层次的站点区域进行划分。对各等级站点区域进行功能定位，并提出多活动集聚区构建方案，实现区域内出行需求的集聚。

再次，提出基于主动引导的公交服务网络分层优化方法。首先基于站点区域对公交出行OD进行集聚分析，识别关键客流并划分等级。进而根据分层优化思路，将公交网络划分为高频快线、干线和支线三个层次，分别以线网总体直达客流密度最大、线网效率最大和可达性最优为目标分别构建了各层线网的优化模型，以站点区域间客流OD为例进行公交线网分层优化设计，得到最终的多层次公交网络，并提出了衔接多层次公交线网的接驳公交服务。

最后，本文从线网规模、网络结构特征、拓扑可达性和客流需求特征多个维度，对优化效果进行定性与定量相结合的评价。

论文以公交服务网络优化方法为研究目标，从居民日常活动-需求角度出发，通过对个体活动空间与公交需求特征的深入分析，构建基于站点活动空间与出行需求耦合的多层次公交服务网络，通过服务资源整合，有效发挥公交集约化运输优势，是促进公共交通发展从“被动适应”向“主动引导”转变的一种积极探索。

关键词：活动-需求，时空特征，站点重要度，站点区域，多层次公交服务网络

Problems such as urban space shortage, traffic congestion, and environmental pollution are becoming increasingly prominent. As a green, economical and intensive motorized transportation mode, public transport has become the key to optimizing land use patterns, alleviating traffic congestion, and realizing sustainable urban transportation development. However, due to the impact of private motorized travel, as well as the low service level and lack of competitiveness of public transport itself, the development of public transport has encountered serious bottlenecks. Faced with the increasingly diversified travel demand and the continuous transformation of public transport travel from “route-level” to “network-level”, the concept of bus route optimization needs to be realized from “passive adaptation” to “active guidance”. Therefore, grasping the characteristics of travelers’ daily activities and public transport demands, exploring the drainage mechanism and grading characteristics of public transport stations, analyzing the topology structure of public transport network, and carrying out network reconstruction and optimization are the keys to innovative public transport service models. In view of the anchoring effect of rail transit on the public transport network, the optimization of the public transport network is mainly considered from the conventional bus in this study.

This study firstly analyzes the distribution characteristics and its generation mechanism of residents’ activities, and the characteristics of bus passenger flow at the bus station level from two dimensions of time and space. First, this study uses the residents’ travel survey data to analyze the travel characteristics; then comprehensively use the mobile phone signaling data and the residents’ travel activity survey data to explore the spatial-temporal distribution characteristics of residents’ daily activities; and use activity origin-destination (OD) as the breakthrough point to analyze the  characteristic changes of activity flow；combined with points of interest (POI) data, this paper analyzes the generation mechanism of activity spatial distribution characteristics from the perspective of land use. And the bus IC card data and route station geographic information data are used to analyzes the spatial-temporal distribution characteristics of bus passenger flow on weekdays and weekends in the study area, and to mine the spatial characteristics of station passenger flow changes.

Secondly, using the complex network theory, the research identifies the key stations and classify the station areas through the station importance evaluation. The existing bus network topology model is constructed, and the structural characteristics of the network are analyzed from multiple perspectives; considering the network structure characteristics, spatial location characteristics and activity demand characteristics of bus stations, a multi-dimensional station importance evaluation index system is established, the weighted Technique for Order Performance by Similarity to an Ideal Solution (TOPSIS) algorithm is applied to identify the key stations in the network. On this basis, according to the station importance score, the K-means clustering algorithm is used to classify the station importance. Based on the evaluation results of station importance and the spatial location characteristics of stations, this paper uses Geohash spatial search algorithm to divide station areas into different levels. Then, the functional positioning of the station areas at each level is carried out, and the construction scheme of multi-activity aggregation areas is proposed to achieve the aggregation of travel demand in the area.

Thirdly, this paper puts forward the method of hierarchical optimization of the public transport service network based on active guidance. Based on the station area, the OD agglomeration analysis of bus passenger flow is first carried out, and the key passenger flow is identified and graded. According to the idea of hierarchical optimization, the network is divided into three levels: high-frequency rapid transit routes, general bus routes and branch bus routes. The optimization model of each hierarchy of route network is constructed respectively, and the optimization object of high-frequency rapid transit routes is to maximum direct passenger flow density of the network, the optimization object of general bus routes is to maximum the network efficiency, and the optimization object of branch bus routes is to optimize accessibility of network. Then taking the passenger flow OD between station areas as an example, the hierarchical optimization design of the bus network is carried out, and the final multi-hierarchy bus network is obtained. And then the connection transport service of multiple-hierarchy public transport network is proposed.

Finally, this paper evaluates the optimization effect qualitatively and quantitatively from multiple dimensions of line network scale, network structure characteristics, topology accessibility and passenger flow demand characteristics.

This study takes the optimization method of public transport service network as the research object, from the perspective of residents’ daily activity travel demand, through in-depth analysis the characteristics of individual activity space and bus travel behavior, a multi-hierarchy public transport service network based on the coupling of station activity space and travel demand is constructed. Through the integration of service resources, the advantages of intensive public transport can be effectively play, which will be an active exploration to promote transformation of public transport development from “passive adaptation” to “active guidance”.

Key Words: activity-demand, spatial-temporal characteristics, station importance, station area, multi-hierarchy public transport service network