Classification of Driver Behavior in Horizontal Curves of Two-Lane Rural Roads
The components of a road directly affect its safety and, by analyzing the consistency of the geometric design, this condition can be assessed. Vehicle speed is a factor in assessing geometric design consistency, and it is considered constant throughout the driver’s transition between the tangents of...
| Main Authors: | , , |
|---|---|
| Format: | Online |
| Language: | eng |
| Published: |
Universidad Pedagógica y Tecnológica de Colombia
2021
|
| Subjects: | |
| Online Access: | https://revistas.uptc.edu.co/index.php/ingenieria/article/view/13410 |
| _version_ | 1801706093743702016 |
|---|---|
| author | Rosas-López, Cristian-David Gaviria-Mendoza, Carlos-Andres Calero-Valenzuela, Carlos-Anibal |
| author_facet | Rosas-López, Cristian-David Gaviria-Mendoza, Carlos-Andres Calero-Valenzuela, Carlos-Anibal |
| author_sort | Rosas-López, Cristian-David |
| collection | OJS |
| description | The components of a road directly affect its safety and, by analyzing the consistency of the geometric design, this condition can be assessed. Vehicle speed is a factor in assessing geometric design consistency, and it is considered constant throughout the driver’s transition between the tangents of horizontal curves. However, in most cases, drivers modify the speed depending on the vehicle’s position before, during, and after the curve. Therefore, a speed profile allows one to describe this behavior better and study the relationship between speed and the curve elements. Previous literature shows that the rural environmental influences the speed of road users.
Additionally, horizontal curve negotiations are linked to a high number of accidents, most of which are caused by driving errors and aggressive/risky driving style. This study aims to determine the optimal number of groups of curve speed profiles that are strongly influenced by the horizontal alignment elements geometry. This research introduces a cluster-based analysis approach to identify speed profiles from field data in two steps; first, the optimal number of speed profiles is determined through the evaluation of several techniques and, second, speed profiles are classified using the k-means method with different sets of data. Free-flow velocity data observed travelling in both directions within 38 horizontal curves on two-lane rural roads in Puerto Rico was used for validation. The validation of the proposed methodology is carried out by directly comparing the number of behaviors observed in users with the geometric and operational characteristics of the curves. The results show that by employing three types of driving styles referred to as Aggressive, Moderate, and Cautious and, adding two inferred variables that consider the time, the driver stays near or above the speed limit and changes acceleration during their movement through the curve, allowing a clearer and more uniform categorization of speed profiles. Furthermore it is observed that the number of grouped profiles have a high dependence on the values taken by the variables of the curve design as expected due to their influence on the perception of safety by drivers. |
| format | Online |
| id | oai:oai.revistas.uptc.edu.co:article-13410 |
| institution | Revista Facultad de Ingeniería |
| language | eng |
| publishDate | 2021 |
| publisher | Universidad Pedagógica y Tecnológica de Colombia |
| record_format | ojs |
| spelling | oai:oai.revistas.uptc.edu.co:article-134102022-06-15T15:51:03Z Classification of Driver Behavior in Horizontal Curves of Two-Lane Rural Roads Clasificación de comportamiento de conductores en curvas horizontales de carreteras rurales de dos carriles Rosas-López, Cristian-David Gaviria-Mendoza, Carlos-Andres Calero-Valenzuela, Carlos-Anibal Perfil de velocidad Clúster Vía de dos carriles Vía rural Curva Diseño geométrico cluster curves geometric design rural road speed profile two-lanes road The components of a road directly affect its safety and, by analyzing the consistency of the geometric design, this condition can be assessed. Vehicle speed is a factor in assessing geometric design consistency, and it is considered constant throughout the driver’s transition between the tangents of horizontal curves. However, in most cases, drivers modify the speed depending on the vehicle’s position before, during, and after the curve. Therefore, a speed profile allows one to describe this behavior better and study the relationship between speed and the curve elements. Previous literature shows that the rural environmental influences the speed of road users. Additionally, horizontal curve negotiations are linked to a high number of accidents, most of which are caused by driving errors and aggressive/risky driving style. This study aims to determine the optimal number of groups of curve speed profiles that are strongly influenced by the horizontal alignment elements geometry. This research introduces a cluster-based analysis approach to identify speed profiles from field data in two steps; first, the optimal number of speed profiles is determined through the evaluation of several techniques and, second, speed profiles are classified using the k-means method with different sets of data. Free-flow velocity data observed travelling in both directions within 38 horizontal curves on two-lane rural roads in Puerto Rico was used for validation. The validation of the proposed methodology is carried out by directly comparing the number of behaviors observed in users with the geometric and operational characteristics of the curves. The results show that by employing three types of driving styles referred to as Aggressive, Moderate, and Cautious and, adding two inferred variables that consider the time, the driver stays near or above the speed limit and changes acceleration during their movement through the curve, allowing a clearer and more uniform categorization of speed profiles. Furthermore it is observed that the number of grouped profiles have a high dependence on the values taken by the variables of the curve design as expected due to their influence on the perception of safety by drivers. Los componentes de una vía infieren directamente en su seguridad vial y, a partir de un análisis de la consistencia en el diseño geométrico se puede cualificar esta condición. La velocidad del vehículo es un factor considerado en la evaluación de la consistencia de diseño geométrico y se considera constante a lo largo de la transición del conductor entre las tangentes de las curvas horizontales. Sin embargo, los conductores en la mayoría de los casos adoptan una velocidad diferente dependiendo de la posición del vehículo antes, durante y después de la curva, por lo cual, un perfil de velocidad permite describir mejor este comportamiento y, estudiar la relación entre la velocidad y los elementos de la curva. La literatura previa muestra que el entorno rural influye en la velocidad de los usuarios de la carretera. Asimismo, las negociaciones en las curvas horizontales están vinculadas a un elevado número de accidentes, la mayoría de los cuales están causados por errores del conductor y un estilo de conducción agresivo/arriesgado. En esta investigación se estudia el número óptimo de comportamientos de los conductores y el proceso de segregación de estos para una base de datos de perfiles de velocidad registrados en campo. Para esto se propone un esquema basado en conglomerados, partiendo de la selección de un numero óptimo de conglomerados aceptable para las características de los datos y posteriormente, utilizando la técnica k-means para agrupar los perfiles semejantes. Se emplea un grupo de datos de velocidad a flujo libre utilizado en 38 curvas horizontales en vías rurales en Puerto Rico que fueron instrumentadas en ambas direcciones. La validación de la metodología propuesta se realiza mediante la comparación directa de la cantidad de comportamientos observados en los usuarios con las características geométricas y operacionales de las curvas. Los resultados muestran que al emplear 3 tipos estilos de conducción referidos como agresivo, moderado y cauto y, agregar dos variables inferidas que consideran el tiempo que el conductor permanece cerca o por encima del límite de velocidad y los cambios en la aceleración durante su desplazamiento por la curva permiten una categorización más clara y uniforme de los perfiles de velocidad. Adicionalmente, se observa que la cantidad de perfiles agrupados presentan una alta dependencia con los valores que toman las variables del diseño de la curva como es de esperarse debido a su influencia en la percepción de la seguridad por parte de los conductores. Universidad Pedagógica y Tecnológica de Colombia 2021-08-12 info:eu-repo/semantics/article info:eu-repo/semantics/publishedVersion application/pdf text/xml https://revistas.uptc.edu.co/index.php/ingenieria/article/view/13410 10.19053/01211129.v30.n57.2021.13410 Revista Facultad de Ingeniería; Vol. 30 No. 57 (2021): July-September 2021 (Continuous Publication); e13410 Revista Facultad de Ingeniería; Vol. 30 Núm. 57 (2021): Julio-Septiembre 2021 (Publicación Continua); e13410 2357-5328 0121-1129 eng https://revistas.uptc.edu.co/index.php/ingenieria/article/view/13410/11059 https://revistas.uptc.edu.co/index.php/ingenieria/article/view/13410/11155 Copyright (c) 2021 Cristian-David Rosas-López, Carlos-Andres Gaviria-Mendoza, Carlos-Anibal Calero-Valenzuela http://creativecommons.org/licenses/by/4.0 |
| spellingShingle | Perfil de velocidad Clúster Vía de dos carriles Vía rural Curva Diseño geométrico cluster curves geometric design rural road speed profile two-lanes road Rosas-López, Cristian-David Gaviria-Mendoza, Carlos-Andres Calero-Valenzuela, Carlos-Anibal Classification of Driver Behavior in Horizontal Curves of Two-Lane Rural Roads |
| title | Classification of Driver Behavior in Horizontal Curves of Two-Lane Rural Roads |
| title_alt | Clasificación de comportamiento de conductores en curvas horizontales de carreteras rurales de dos carriles |
| title_full | Classification of Driver Behavior in Horizontal Curves of Two-Lane Rural Roads |
| title_fullStr | Classification of Driver Behavior in Horizontal Curves of Two-Lane Rural Roads |
| title_full_unstemmed | Classification of Driver Behavior in Horizontal Curves of Two-Lane Rural Roads |
| title_short | Classification of Driver Behavior in Horizontal Curves of Two-Lane Rural Roads |
| title_sort | classification of driver behavior in horizontal curves of two lane rural roads |
| topic | Perfil de velocidad Clúster Vía de dos carriles Vía rural Curva Diseño geométrico cluster curves geometric design rural road speed profile two-lanes road |
| topic_facet | Perfil de velocidad Clúster Vía de dos carriles Vía rural Curva Diseño geométrico cluster curves geometric design rural road speed profile two-lanes road |
| url | https://revistas.uptc.edu.co/index.php/ingenieria/article/view/13410 |
| work_keys_str_mv | AT rosaslopezcristiandavid classificationofdriverbehaviorinhorizontalcurvesoftwolaneruralroads AT gaviriamendozacarlosandres classificationofdriverbehaviorinhorizontalcurvesoftwolaneruralroads AT calerovalenzuelacarlosanibal classificationofdriverbehaviorinhorizontalcurvesoftwolaneruralroads AT rosaslopezcristiandavid clasificaciondecomportamientodeconductoresencurvashorizontalesdecarreterasruralesdedoscarriles AT gaviriamendozacarlosandres clasificaciondecomportamientodeconductoresencurvashorizontalesdecarreterasruralesdedoscarriles AT calerovalenzuelacarlosanibal clasificaciondecomportamientodeconductoresencurvashorizontalesdecarreterasruralesdedoscarriles |
