| Abstrakt: |
Current research predominantly relies on single travel data sources to simulate single-modal transportation networks, making it challenging to objectively and comprehensively reflect the real travel characteristics of residents. This paper presents a modeling method of multi-modal transportation network and a multifeature fusion algorithm for analyzing resident travel characteristics, taking Shenzhen as a case study. Firstly, taking into consideration the spatio temporal variations in transportation networks, a partitioning algorithm of multiscale transportation zones is designed to simplify the complex network. Next, a multi-model transportation network is constructed by connecting station path attributes with transportation zones. Finally, mobile trajectory mining and feature fusion techniques are adopted to design a fusion algorithm of multi-modal travel features, systematically depicting resident travel characteristics in a multi-model context. Experimental results demonstrate what compared to using a single data source alone, the proposed method effectively mitigates data bases, provides a more accurate representation of traffic density among different transportation regions, and offers a more comprehensive insight into residents' travel behaviors and preferences, and can better provide scientific decision-making for urban planning and transportation management. [ABSTRACT FROM AUTHOR] |
