消除公交串车——一种基于GPS与GIS的实时控制方法

城市公共交通系统是一个包含人、车、路、环境、能量和信息的高度非线性的大规模复杂系统。公交串车问题严重降低了公交服务的可靠性，本文综合运用系统工程理论与方法，给出了公交串车与大间隔的定义，提出了公交串车与大间隔问题的成因分析模型、基于BP人工神经网络的行车间隔预测模型和基于GPS与GIS的实时动态反馈控制模型。最后，对北京公交300路车所采集的680组数据进行了实证分析。分析结果表明所使用的方法能有效地消除公交串车问题，同时也能为公交企业自身的运营管理和乘客的出行规划提供有价值的决策参考。

Bus Bunching Elimination—A GPS/GIS-based Real-time Control Approach

Urban public transport system is a highly nonlinear and large-scale complex system, involving human beings, vehicles, roads, environment, energy and information. Bus bunching problem seriously reduces the reliability of bus service. To solve this problem, theories and methods of systems engineering are used in this research. Firstly, definitions of bus bunching and long headway are given. Then the mathematical analysis model of the bus bunching and long headway problems, the BP-artificial neural network （BP-ANN）-based bus headway prediction model and the GPS/GIS-based real-time dynamic feedback control model are given, respectively. Finally, a case study of 680 groups of data collected from the line 300 of Beijing public transport is presented. The results indicate that the bus bunching problem can be effectively eliminated and also it can provide valuable decision-making references for both the operation of public transport companies and the travel planning of public transport passengers.