Design and implementation of traffic flow simulation
Traffic flow simulation has become important in recent years. It is an alternative to evaluate traffic conditions of various transportation infrastructure scenarios. There is a wide range of models that try to describe vehicle movement in a traffic stream. However, microscopic models have prevailed...
Guardat en:
| Autor principal: | |
|---|---|
| Format: | bachelorThesis |
| Idioma: | eng |
| Publicat: |
2021
|
| Matèries: | |
| Accés en línia: | http://repositorio.yachaytech.edu.ec/handle/123456789/372 |
| Etiquetes: |
Afegir etiqueta
Sense etiquetes, Sigues el primer a etiquetar aquest registre!
|
Afegir etiqueta | Sumari: | Traffic flow simulation has become important in recent years. It is an alternative to evaluate traffic conditions of various transportation infrastructure scenarios. There is a wide range of models that try to describe vehicle movement in a traffic stream. However, microscopic models have prevailed as the most used in the field of simulation, specially car-following models. Commonly traffic simulators are made up of three relationships: car-following, lane changing, and gap acceptance. Car-following models can reproduce various microscopic traffic phenomena. These models are the most important representatives among currently used microscopic simulators. These models are based on the idea that vehicles adjust their parameters (e.g., speed, headway) depending on the preceding vehicle. On the other hand, each vehicle in the simulation will freely accelerate until it reaches the desired speed when it is not blocked by a vehicle in front. The model studied in this work is the Gipps' car-following model which was developed back in 1981. Although several models have been developed since then, the Gipps' model remains one of the most used and relevant models today. The original model and its modifications continue to be the nucleus of several academic and commercial traffic simulators. It can also cooperate with the Gipps' lane change model or other lane change models to build more complex traffic simulators. Also, Gipps' car-following model can imitate real traffic flow characteristics such as heterogeneity of traffic, traffic congestion, and collision avoidance. For this reason, we consider it important to carry out an implementation and evaluation of this model to update results from previous works. Our implementation of the model was two-dimensional and done in the Python programming language. The experiments presented in this work show that our implementation of the model behaves like the original Gipps' model. Also, the experiments show that we can replicate various traffic phenomena, such as traffic congestion and heterogeneous traffic. Therefore, our implementation of the model can be used in future works as a reliable model for traffic simulation. |
|---|