Multi-Class Vehicle Segregation for Enhanced Safety and Efficiency of Mixed Traffic Networks
| Autor: | Aathira K. Das, Bhargava Rama Chilukuri |
|---|---|
| Jazyk: | angličtina |
| Rok vydání: | 2024 |
| Předmět: | |
| Zdroj: | IEEE Access, Vol 12, Pp 116368-116383 (2024) |
| Druh dokumentu: | article |
| ISSN: | 2169-3536 26936402 |
| DOI: | 10.1109/ACCESS.2024.3447049 |
| Popis: | The heterogeneity of vehicle types and their lane-free movement are the unique characteristics of mixed traffic conditions. Modelling and controlling such a traffic system is challenging. Towards this, to overcome the adverse effects of heterogeneity, segregation of traffic based on vehicle type is the viable solution proposed in this study. Two mathematical formulations are developed for multi-class segregation problems to identify a vehicle type or a combination of vehicle types to be segregated from the remaining traffic to enhance efficiency and safety in mixed-traffic networks. The corresponding objectives are related to minimising the total system travel time and network crash risk. The developed models were evaluated using a real-world mixed traffic network case study and it was observed that the total system travel time savings were 10% compared to the case of conventional traffic assignment. Sensitivity analysis shows that the segregation patterns are repeatable across compositions, demands, and parameter values. Based on the observations, a heuristic solution is proposed for the multi-class vehicle segregation problem. The heuristic solution involves routing different vehicle types in the network, and the evaluation results showed that the heuristic method could provide a good trade-off solution with respect to efficiency and safety. Considering the extreme point solutions of the two single-objective formulations, the total system travel time value is 11% higher than the single objective case of travel time and the total crash risk is 90% higher than the single objective case of crash risk. The findings in this paper are expected to improve the modelling and traffic assignment strategies for mixed traffic networks by addressing their unique characteristics. |
| Databáze: | Directory of Open Access Journals |
| Externí odkaz: |
|