数据驱动的公交网络动态优化调整方法

Data-driven Optimization Methods for Dynamic Transit Network Adjustment

为了更好地满足不断变化的公交出行需求和提高公交运营效率,本文提出一种数据驱动的公交网络动态优化调整方法——滚动决策,频率为先,增删迭代.该方法由两部分组成:第1部分是数据驱动的公交网络优化调整时机决策;第2部分是公交网络优化调整措施决策,包括发车频率优化和线网结构优化.该方法的具体思路为:对于给定的现状公交网络,通过数据建模分析,动态判断公交网络优化调整的触发时机,若触发优化,则优先进行发车频率优化调整,如果发车频率优化调整不能满足预期系统目标,则进行线网结构优化调整,线网结构优化调整策略简化为新增线路和删除线路的迭代.数值实验表明,该方法能比较充分利用智能公交系统采集的数据,可操作性强,可为公交网络动态优化提供决策参考.

In order to meet the variable demand of bus transit and improve the efficiency of operation, this paper proposes a data-driven dynamic transit network optimization method-rolling decision, frequency priority, iterating of generating and deleting routes. The method consists of two parts： the first part is the decision making of data-driven optimization timing; the second part is the decision making of optimization adjustment measures, which including the departure frequency optimization and the bus network structure optimization. The specific idea of this method is that through data modeling and analysis on the current transit network, the system determines whether to take optimization adjustment of transit network on time. If the network needs to be optimized, the system optimizes the frequency of transit lines with priority. If not, it should optimize the structure of the transit network by iterating generating and deleting routes. The example analysis shows that this method can make full use of the data collected by the Intelligent Public Transit System, and the operation is convenient. Finally, it can provide decision reference for dynamic transit network optimization.