Data-driven traffic zone division in smart city: Framework and technology

Abstract

The development of smart city and smart mobility has put forward requirements for the traffic zone division. The traditional traffic zone division is limited by data with low update frequency and low accuracy, which can hardly meet the needs. With the progress of ICTs, data-driven traffic zone division is developing gradually. Many scholars have tried to divide traffic zone through GPS and CDR. However, there are still shortcomings in the data-driven traffic zone division. On the one hand, different scholars use different data sources and division methods, and there is no unified framework that can describe different ways. On the other hand, many technical problems such as unsatisfactory zone boundary, single data source and clustering algorithm only considering similarity have not been solved well. Therefore, this paper first proposes a technical framework for data-driven traffic zone division, which describes the different methods uniformly by dividing the data-driven traffic zone division into four parts. Based on this framework, this paper proposes a traffic zone division method based on CSD, which solves the above technical problems successfully, and finally is applied in the traffic zone division of Foshan, China.