Scholarlink

<p>In recent decades, the simulation of vehicular traffic at roundabouts has received notable attention among traffic professionals. Developing reliable tools, specifically to model roundabout traffic, is vital given the growing interest in analyzing roundabout traffic operation and safety. These tools can play an important role in anticipating the effectiveness of any traffic design and control change at roundabouts. The social force-based model (SFM) has been widely used in simulating the movements of road users and their interactions at different traditional and shared space zones. Despite their strengths, SFMs have not been applied to simulate the movements and interactions of vehicles at traditional or even shared-space roundabouts. This study proposes an SFM for the movement of mixed vehicular traffic at traditional roundabouts. The calibration and validation of the model were undertaken using open trajectory datasets from a roundabout in Germany. The local transferability of the model was tested using datasets from other locations in Germany. The results showed that the calibrated model could replicate different traffic movements at the roundabout. Moreover, the calibrated model showed good performance in terms of the root mean square error (RMSE) of predicted vehicle trajectories and speeds, final destination error (FDE), and average distance error (ADE) when applied at other locations in Germany. The RMSE of trajectory and speed, ADE, and FDE from using the validation dataset were 0.35&#xa0;m, 0.19&#xa0;m/sec, 0.26&#xa0;m, and 0.57&#xa0;m, respectively. The external transferability of the model to another environment in a different country was further explored where the model parameters were re-calibrated using trajectory data from a roundabout in the USA. From the comparison between the calibrated values of each parameter, it was concluded that five of the eleven common calibrated parameters are transferable, while the other parameters should be re-calibrated to improve the performance of the model at different locations. Finally, it was concluded that the proposed model can depict the movement of different vehicles at the roundabout after proper calibration for the non-transferable parameters.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Development of a Social Force Simulation Model for Mixed Vehicular Traffic at Roundabouts

  • 期刊论文

  • 发表日期：

    2025年10月29日

查看全文
  • Salma Abo-Bakr,
  • Karim Ismail,
  • Mohamed El Esawey,
  • Mostafa H. Tawfeek

摘要

In recent decades, the simulation of vehicular traffic at roundabouts has received notable attention among traffic professionals. Developing reliable tools, specifically to model roundabout traffic, is vital given the growing interest in analyzing roundabout traffic operation and safety. These tools can play an important role in anticipating the effectiveness of any traffic design and control change at roundabouts. The social force-based model (SFM) has been widely used in simulating the movements of road users and their interactions at different traditional and shared space zones. Despite their strengths, SFMs have not been applied to simulate the movements and interactions of vehicles at traditional or even shared-space roundabouts. This study proposes an SFM for the movement of mixed vehicular traffic at traditional roundabouts. The calibration and validation of the model were undertaken using open trajectory datasets from a roundabout in Germany. The local transferability of the model was tested using datasets from other locations in Germany. The results showed that the calibrated model could replicate different traffic movements at the roundabout. Moreover, the calibrated model showed good performance in terms of the root mean square error (RMSE) of predicted vehicle trajectories and speeds, final destination error (FDE), and average distance error (ADE) when applied at other locations in Germany. The RMSE of trajectory and speed, ADE, and FDE from using the validation dataset were 0.35 m, 0.19 m/sec, 0.26 m, and 0.57 m, respectively. The external transferability of the model to another environment in a different country was further explored where the model parameters were re-calibrated using trajectory data from a roundabout in the USA. From the comparison between the calibrated values of each parameter, it was concluded that five of the eleven common calibrated parameters are transferable, while the other parameters should be re-calibrated to improve the performance of the model at different locations. Finally, it was concluded that the proposed model can depict the movement of different vehicles at the roundabout after proper calibration for the non-transferable parameters.