Multiclass Cell Transmission Model for Heterogeneous Mobility in General Topology of Road Network

Abstract

This article is concerned with a mathematical generalization of macroscopic road network mobility model, the cell transmission model. The objective is to enhance the conventional single-class cell transmission model (S-CTM) to a more generalized multiclass cell transmission model (M-CTM). This latter model can take into account the mixed composition of vehicle classes (i.e., with distinct free-flow speed and vehicle length). The aim is to introduce head-of-cell and end-of-cell vehicles into the model. Cascading, merging, and diverging scenarios of cells for M-CTM have been formulated so that the model can be applied to a general topology of road network. Experimental results are provided to compare S-CTM and M-CTM in 6 network test cases with the microscopic simulator—MITSIM—and to actual measurement of network traffic conditions. The results suggest that M-CTM is significantly more accurate than S-CTM in the tested networks with nonstationary vehicle composition, with platoon dispersions and with time-varying jam status. In addition, accuracy improvement of M-CTM is obtained without compromising the computational complexity. It is therefore expected that the proposed M-CTM would be well applicable in practice.

Keywords:

• cell transmission model
• heterogeneous mobility
• macroscopic model
• traffic network

The authors thank the Thailand Research Fund under Grant No. MRG4880172 for sponsoring this work, the Bangkok Metropolitan Administration for traffic data, and the National Electronics and Computer Technology Center, Thailand for the support on the Intelligent Area Traffic Control and Real-Time Traffic Information System Development Project, and the Centre of Excellence in Lightwave and High-Speed Communications, TRIDI. Also, special thanks are given to Dr. S. Narupiti and Mr. P. Komolkiti for their help with conducting the experiment.