Publication: Eco-driving: ahorro de energía basado en el comportamiento del conductor
Loading...
Identifiers
Publication date
2014
Defense date
2014-06-13
Authors
Advisors
Tutors
Journal Title
Journal ISSN
Volume Title
Publisher
Impact
Export
Abstract
El crecimiento del número de vehículos en circulación ha experimentado un fuerte aumento en los últimos 20 años. La generalización del uso del automóvil ha tenido efectos muy positivos en la economía de los países. Sin embargo, también ha provocado grandes problemas debido la contaminación y a la cantidad de energía que consumen. Por otra parte, la mayoría de los vehículos emplean hidrocarburos, que no se encuentran disponibles en todas las regiones, provocando dependencias energéticas entre países. Además, su extracción tiene un impacto muy grande en el medioambiente.
Los vehículos se han convertido en un problema importante para los gobiernos y los habitantes, que sufren enfermedades respiratorias provocadas por los gases que emiten. Ante estos inconvenientes, los gobiernos han desarrollado normativas para regular las emisiones de los vehículos. Las conductores también han empezado a exigir vehículos que consuman menos debido al aumento del precio del combustible, convirtiéndose en un factor muy importante a la hora de comprar un vehículo. Todo esto ha contribuido a que los fabricantes introduzcan en los vehículos mejoras orientadas a reducir el consumo de combustible como: optimización del motor, reducción del peso del vehículo, motores híbridos, mejoras aerodinámicas. Sin embargo, estas medidas resultan insuficientes porque el porcentaje de vehículos antiguos en circulación es muy alto.
El consumo de combustible depende del vehículo, el entorno y el comportamiento del conductor. Recientemente ha surgido un método para reducir el consumo denominado “Eco-Driving”. Esta técnica de conducción se basa en el control óptimo de las variables que controla el conductor como la velocidad, la aceleración, la desaceleración y la marcha. El ahorro se consigue al minimizar las pérdidas de energía. Este método para mejorar el consumo y reducir la emisión de gases contaminantes es muy útil porque es independiente de la tecnología del vehículo. Sin embargo, el conductor necesita contar con conocimientos sobre conducción eficiente.
Esta tesis se centra en la necesidad de que los conductores aprendan cómo conducir eficientemente siguiendo unas normas, basadas en la física, que reducen las demandas innecesarias de energías. Para ello se propone el uso de un asistente que evalúa el comportamiento del conductor desde el punto de vista de la eficiencia, y le propone mejoras. Además, a diferencia de otras propuestas, el sistema proporciona información sobre el entorno cercano para que el usuario pueda tomar decisiones con la suficiente antelación. Esta información es muy importante porque la clave del “eco-driving” es la capacidad del conductor para predecir el estado de la carretera en un futuro cercano. Otro de los objetivos principales de este trabajo de investigación es desarrollar métodos para motivar al usuario a conducir eficientemente, ya que en muchos estudios previos se demuestra que el conductor tiende a volver a sus hábitos de conducción previos a pesar de haber recibido formación sobre “eco-driving”.
Los resultados obtenidos en la etapa de experimentación muestran un ahorro de hasta el 22.5% de combustible utilizando la propuesta descrita en esta tesis. No obstante, este porcentaje depende de la respuesta del usuario a las recomendaciones y de su estilo de conducción previo. Además se observó que es importante conocer el perfil de conducción del usuario para ajustar los consejos, y que el aprendizaje se haga de forma progresiva. En caso contrario, el conductor perderá el interés. Conocer el tipo de usuario también es relevante para elegir el incentivo que se usará como retroalimentación. Finalmente, se constata que es imprescindible emplear métodos para motivar al usuario a conducir eficientemente, siendo la gamificación una buena estrategia para conseguirlo.
The growth in the number of vehicles in circulation has experienced a strong increase in the last 20 years. The widespread use of the automobile has had very positive effects on the economy of the countries. However, they have also led to major problems due to the pollution produced and the amount of energy required. On the other hand, most of the vehicles employ hydrocarbons, which are not available in all regions, causing energy dependencies between countries. In addition, its extraction has a very large impact on the environment. The vehicles have become an important problem for governments and residents who suffer from respiratory diseases caused by greenhouses emissions. In order to avoid these problems, governments have developed regulations to control the emissions from vehicles. The drivers have also begun to require efficient vehicles due to the increase in fuel prices. Currently, the fuel consumption is a very important factor to buy a vehicle. All this has caused manufacturers to introduce improvements in vehicles in order to minimize the fuel consumption such as: engine optimization, vehicle weight reduction, hybrid engines, aerodynamic improvements. However, these measures are insufficient because the percentage of older vehicles is still very high. Fuel consumption depends on the vehicle, the environment and the behavior of the driver. Recently, a method called "Eco-Driving" has gained popularity. This driving allows us to save fuel. It is based on the optimization of the parameters that the user controls such as: speed, acceleration, deceleration and gear. Fuel saving is achieved by minimizing the energy losses. This method to improve fuel consumption and reduce the emission of greenhouse gases is very useful because it is independent of the vehicle technology. However, the driver needs to know about efficient driving rules. This thesis is focused on the need to learn how to drive efficiently following a set of rules, based on physics, which reduce unnecessary energy demands. In this work, we propose the use of an assistant who evaluates the driver behavior from the point of view of efficiency, and recommends improvements in order to save fuel. Furthermore, unlike other proposals, the system provides information about the nearby environment. Therefore, the user can take decisions in advance. This information is very important because the key to "eco-driving" is the driver's ability to predict the road state in the near future. The other main objective is to develop methods to motivate the user for driving efficiently, since many previous studies demonstrates that the driver tends to return to their previous driving habits, even after having received training on "eco-driving". The results of the experimentation show a saving of up to 22.5% of fuel using the proposal described in this thesis. However, this percentage depends on user´s response to the recommendations and their previous driving style. We also highlight that is essential to know the driving profile in order to adjust the eco-driving tips, and make the learning progressively. Otherwise, the driver will lose interest. In addition, understanding the user type is relevant to choose the type of feedback. Finally, the work in this thesis demonstrates that it is essential to use methods for motivating the user. Gamification is a good strategy to get it.
The growth in the number of vehicles in circulation has experienced a strong increase in the last 20 years. The widespread use of the automobile has had very positive effects on the economy of the countries. However, they have also led to major problems due to the pollution produced and the amount of energy required. On the other hand, most of the vehicles employ hydrocarbons, which are not available in all regions, causing energy dependencies between countries. In addition, its extraction has a very large impact on the environment. The vehicles have become an important problem for governments and residents who suffer from respiratory diseases caused by greenhouses emissions. In order to avoid these problems, governments have developed regulations to control the emissions from vehicles. The drivers have also begun to require efficient vehicles due to the increase in fuel prices. Currently, the fuel consumption is a very important factor to buy a vehicle. All this has caused manufacturers to introduce improvements in vehicles in order to minimize the fuel consumption such as: engine optimization, vehicle weight reduction, hybrid engines, aerodynamic improvements. However, these measures are insufficient because the percentage of older vehicles is still very high. Fuel consumption depends on the vehicle, the environment and the behavior of the driver. Recently, a method called "Eco-Driving" has gained popularity. This driving allows us to save fuel. It is based on the optimization of the parameters that the user controls such as: speed, acceleration, deceleration and gear. Fuel saving is achieved by minimizing the energy losses. This method to improve fuel consumption and reduce the emission of greenhouse gases is very useful because it is independent of the vehicle technology. However, the driver needs to know about efficient driving rules. This thesis is focused on the need to learn how to drive efficiently following a set of rules, based on physics, which reduce unnecessary energy demands. In this work, we propose the use of an assistant who evaluates the driver behavior from the point of view of efficiency, and recommends improvements in order to save fuel. Furthermore, unlike other proposals, the system provides information about the nearby environment. Therefore, the user can take decisions in advance. This information is very important because the key to "eco-driving" is the driver's ability to predict the road state in the near future. The other main objective is to develop methods to motivate the user for driving efficiently, since many previous studies demonstrates that the driver tends to return to their previous driving habits, even after having received training on "eco-driving". The results of the experimentation show a saving of up to 22.5% of fuel using the proposal described in this thesis. However, this percentage depends on user´s response to the recommendations and their previous driving style. We also highlight that is essential to know the driving profile in order to adjust the eco-driving tips, and make the learning progressively. Otherwise, the driver will lose interest. In addition, understanding the user type is relevant to choose the type of feedback. Finally, the work in this thesis demonstrates that it is essential to use methods for motivating the user. Gamification is a good strategy to get it.
Description
Keywords
Eco-driving, Ahorro energético, Optimización del consumo de combustible, Sistemas inteligentes de transporte, Gamificación, Lógica difusa, Clustering, Inteligencia artificial