Dynamic Bus Scheduling For Single Line Under The Internet Of Things Environment

In recent years, urban traffic congestion is getting worse and worse with the rapid acceleration of urbanization and the fast increment of vehicle population in China. Developing urban public transport is a constructive way to mitigate traffic congestion. Improving the service quality and the attractiveness of public transportation by applying different bus scheduling strategies is an important field in the public transportation research. In practical operation process, static bus planning probably can’t meet the needs of bus operators, thus the operators have to implement dynamic scheduling to keep the operation stay efficient and to provide high quality services. With the development of Internet of Things (IOT) and Intelligent Transport System (ITS), how to use mass real-time information to achieve effective real-time control has become a new highlighted problem of the dynamic scheduling research. Based on the above problem, research of dynamic bus scheduling under the IOT environment has been conducted in this thesis, and the main work includes the following aspects:First, this thesis proposes an urban dynamic bus scheduling framework model. Then, a dynamic bus scheduling flow chart based on passenger flow coefficient is presented. Control strategies such as dynamic bus dispatching, expressing, inter-zone vehicle, short-turning and part-way deadheading are considered in this flow chart.Second, a dynamic bus dispatching model for a single line based on rolling horizon framework is presented and formulated based on mass real-time information under the IOT environment. The objective of this model is to minimize the total passenger waiting time and the genetic algorithm (GA) is developed to solve this optimization model. Some typical experiments are conducted, and the results show that the performance of the dynamic bus dispatching strategy meets the expectation. Then a dynamic evolution algorithm is also developed in order to improve the performance of the dynamic bus dispatching strategy in the multi-horizon case and the results show that dynamic evolution algorithm can achieve better efficiency than basic GA.Third, station-skipping models for both one operating direction and two operating directions are established for a single bus line. The one-direction model refers to the station-skipping control strategies which include expressing and inter-zone vehicle and the double-direction model refers to the control strategies which include short-turning and part-way deadheading. Mass traffic real-time information under the IOT environment is considered in the formulation of the model. Then simulation experiments are conducted based on the dynamic bus scheduling flow chart, and the obtained result shows that the system cost can be reduced by 32%.Finally, an intelligent urban bus scheduling software system is developed based on the proposed dynamic bus scheduling methods. The dynamic scheduling module and the transit simulation module which embedded the dynamic bus scheduling theory are developed by using with C# language based on Microsoft Visual Studio 2010 development platform and SQL server 2008. By combining with the transit network data, the intelligent urban bus scheduling platform software system can provide variety of functions for dynamic scheduling.