基于柔性的城市公交系统网络运能配置研究

To the variability, dynamics and uncertainty of travel demand for urban transit network, the transit systems with flexibility for urban traffic systems should be established to improve service quality of transit, urgently. In this program, the models and methods to the capacity flexibility of transportation network for urban transit systems are proposed by the theory of systems analysis and management flexibility. Then the uncertain programming models and algorithms to the capacity flexibility of transportation network for urban transit systems are established in terms of uncertainty of OD demands. The capacity allocation and uncertain programming models of transportation network for urban transit systems with flexibility are extended on this base. It reflects the line capacity allocation of transportation network for urban transit systems. The algorithms with stochastic simulation, local search and Meta-heuristics are designed for improving performance and efficiency of algorithms. Moreover, the simulation models for transit transfer systems with Petri nets theory and computer simulation technique is presented according to the optimal objective of the capacity flexibility. And it is applied to the transfer design of the station for high speed railway and transit transfer station. This reflects the node capacity allocation of transportation network for urban transit systems. At last, the basic data are obtained by transit surveys in experimental cities. The diversification and uncertainty of time and space to OD travel demand for urban transit is analyzed quantitatively with Geographic Information System (GIS) software. The feasibility and the effectiveness of models and algorithms above mentioned is examined by experimental cities. The improvement recommendations for the transit problem in experimental cities are presented with the achievement of theory research. This could provide decision support for the management of transit operation and the design of transit transfer system in experimental cities.

为了应对城市公交出行需求时空分布的多样性、动态性及不确定性，城市交通迫切需要建立柔性的公交系统，改善公交服务质量。本项目主要运用系统科学和柔性理论，提出城市公交系统网络运能柔性模型、不确定性规划模型和度量方法。在此基础上对模型进行扩展，建立基于柔性的城市公交系统网络运能配置模型和不确定性规划模型，体现运能在“线”上的配置，运用随机模拟技术、邻域搜索技术和Meta-heuristics算法进行算法设计，提升算法的性能和效率。运用Petri网理论建立基于Petri网的公交换乘枢纽仿真模型，结合我国高铁站点和城市公交换乘系统设计进行应用分析，体现运能在“点”上的配置。通过典型城市收集必要的公交基础数据，以地理信息系统(GIS)软件为平台，定量分析公交OD出行需求时空变化及其不确性，检验上述理论模型和方法的可行性和有效性并提出改进建议，为典型城市公交运营组织管理和公交换乘系统布局设计提供决策支持。

为了应对城市公交出行需求时空分布的多样性、动态性及不确定性，城市交通迫切需要建立柔性的公交系统，改善公交服务质量。本项目首先考虑了公交发车频率、车辆能力、车辆响应时间、公交运营组织等因素，构建城市公交系统网络运能柔性模型，并结合公交系统网络运能柔性模型特点，设计优化算法；再进一步构建基于柔性的城市公交系统网络运能优化配置模型和优化算法，并以贵阳市为案例，分析各因素通过自身调整响应需求变化的能力，说明各因素间及运能柔性间的相互作用机制。其次，应用Memetic算法优化城市公交网络运能，主要是优化公交路径布局和公交服务频率。所提出的解算法通过编制两类线路集列表，及设计四类邻域搜索操作：2-opt move（Type A），2-opt move（Type B），swap move和relocation move，获得更好的遗传算法（GA）染色体。并通过有效的试错步骤修正邻域搜索操作，提升搜索效率。该优化算法通过benchmark问题进行实证研究，与传统算法进行比较，在计算性能上得到明显改善，结合贵阳市公交线网实际，给出贵阳市公交线网接驳的优化方案。然后，利用贵阳市出租车GPS数据，建立ArcGIS空间模型计算路段平均行程速度，分析道路拥堵状态的时空演化规律，通过划分交通小区计算出各个交通小区平均行程速度和空间自相关Moran’s I指数，对贵阳市中心城区交通状态的空间自相关性和演化特征进行分析。最后，结合ArcGIS软件平台，构建了城市常规公交服务网络覆盖度计算模型，根据贵阳市中心城区的商业网点POI数据，在对贵阳市中心城区的各类商业网点的空间分布特征分析基础上，对常规公交服务网络与商业网点空间匹配关系开展进一步研究。此外，基于ArcGIS的空间分析，分别对贵阳市中心城区道路网络的可达性和城市大型商业建筑空间分布格局进行测算和比较，还分析了中心城区路网可达性与大型商业建筑空间分布的发展现状及两者间的相互关系。这些研究的开展，不仅对丰富和发展公交柔性理论具有重要的理论价值，而且对提高贵阳市公交系统资源有效利用和改善公交服务质量也具有重要的现实意义。