Publicación: Modelo de bajo costo para la estimación de emisiones contaminantes basado en GPS y aprendizaje automático
No hay miniatura disponible
Archivos
Fecha
2022
Editor/a
Director/a
Tutor/a
Coordinador/a
Prologuista
Revisor/a
Ilustrador/a
Derechos de acceso
info:eu-repo/semantics/openAccess
Título de la revista
ISSN de la revista
Título del volumen
Editor
Universidad Nacional de Educación a Distancia (España), Universidad Politécnica de Madrid. Departamento de Ingeniería Mecánica
Resumen
El presente trabajo presenta un método novedoso para la estimación de los contaminantes emitidos por vehículos propulsados con motores de combustión interna en conducciones reales de conducción, sin la necesidad de campañas extensas de medición ni el uso de instrumentación en el vehículo por períodos prolongados de tiempo; para lo cual se basa en las señales de posicionamiento y velocidad generadas por el GPS (Global Positioning System) y la aplicación de aprendizaje automático. Para la obtención de los datos de entrenamiento y validación del modelo se realizan dos pruebas en ruta mediante las directivas de Euro 6 para la estimación de contaminantes mediante RDE (Real Driving Emissions), en las que se utiliza un sistema de medición de emisiones portátil, y un registrador que almacena los datos provenientes de OBD (On Board Diagnostics) y del GPS. A partir de los datos obtenidos en la primera ruta se determinan las prestaciones del vehículo y mediante aprendizaje automático se genera el modelo que estima las emisiones contaminantes, el cual es validado con los datos de la segunda ruta. Al comparar los resultados genera-dos por el modelo frente a los medidos en el RDE se obtienen errores relativos (%) de 0.0976, -0.2187, 0.2249 y -0.1379 en los factores de emisión de CO2, CO, HC y NOx respectivamente. Finalmente se alimenta al modelo con datos obtenidos en 1218 km de conducción aleatoria, obteniendo similares resultados a modelos basados en OBD y más próximos a las condiciones reales de circulación que generan modelos como el IVE (International Vehicle Emissions). La obtención de datos mediante el OBD y GPS presentes en los vehículos actuales es económica y la aplicación de modelos de aprendizaje automático para la estimación de emisiones contaminan-tes es una opción que abre un campo de trabajo sin la necesidad de contar con equipos a bordo de adquisición más costosos para la realización de campañas experimentales en muestras limitadas de vehículos por los costes asociados.
This paper presents a novel method for estimating the pollutants emitted by vehicles powered by internal combustion engines in real driving conditions, without the need for extensive measurement campaigns or the use of instrumentation in the vehicle for prolonged periods of time; for which it is based on the positioning and speed signals generated by the GPS (Global Positioning System) and the machine learning application. To obtain the data for training and validation of the model, two road tests are carried out using the Euro 6 directives for the estimation of pollutants through RDE (Real Driving Emissions), in which a portable emissions measurement system is used. and a logger that stores data from OBD (On Board Diagnostics) and GPS. From the data obtained in the first route, the performance of the vehicle is determined and through automatic learning, the model that estimates the polluting emissions is generated, which is validated with the data of the second route. When comparing the results generated by the model against those measured in the RDE, relative errors (%) of 0.0976, -0.2187, 0.2249 and -0.1379 are obtained in the emission factors of CO2, CO, HC and NOx, respectively. Finally, the model is fed with data obtained in 1218 km of random driving, obtaining similar results to OBD-based models and closer to the real traffic conditions generated by models such as IVE (International Vehicle Emissions). Obtaining data through the OBD and GPS present in current vehicles is cheap and the application of automatic learning models for the estimation of polluting emissions is an option that opens up a field of work without the need for on-board equipment. more expensive acquisition costs to carry out experimental campaigns on limited samples of vehicles due to the associated costs.
This paper presents a novel method for estimating the pollutants emitted by vehicles powered by internal combustion engines in real driving conditions, without the need for extensive measurement campaigns or the use of instrumentation in the vehicle for prolonged periods of time; for which it is based on the positioning and speed signals generated by the GPS (Global Positioning System) and the machine learning application. To obtain the data for training and validation of the model, two road tests are carried out using the Euro 6 directives for the estimation of pollutants through RDE (Real Driving Emissions), in which a portable emissions measurement system is used. and a logger that stores data from OBD (On Board Diagnostics) and GPS. From the data obtained in the first route, the performance of the vehicle is determined and through automatic learning, the model that estimates the polluting emissions is generated, which is validated with the data of the second route. When comparing the results generated by the model against those measured in the RDE, relative errors (%) of 0.0976, -0.2187, 0.2249 and -0.1379 are obtained in the emission factors of CO2, CO, HC and NOx, respectively. Finally, the model is fed with data obtained in 1218 km of random driving, obtaining similar results to OBD-based models and closer to the real traffic conditions generated by models such as IVE (International Vehicle Emissions). Obtaining data through the OBD and GPS present in current vehicles is cheap and the application of automatic learning models for the estimation of polluting emissions is an option that opens up a field of work without the need for on-board equipment. more expensive acquisition costs to carry out experimental campaigns on limited samples of vehicles due to the associated costs.
Descripción
Categorías UNESCO
Palabras clave
modelo bajo costo, estimación de contaminantes, aprendizaje automático, emisiones basadas en GPS
Citación
Centro
E.T.S. de Ingenieros Industriales
Departamento
Mecánica