Publicación: Técnicas de aprendizaje automático para el control de depuradoras. Resolución del problema del Nitrógeno
Fecha
2018-06-26
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). Escuela Técnica Superior de Ingeniería Informática. Departamento de Inteligencia Artificial
Resumen
En 2015, la ONU aprobó la Agenda 2030 sobre el Desarrollo Sostenible, un conjunto de objetivos globales para erradicar la pobreza, proteger el planeta y asegurar la prosperidad para todos. El Objetivo 6 consiste en garantizar la disponibilidad de agua y su gestión sostenible y el saneamiento para todos. El día Mundial del Agua 2017 se dedicó a la importancia del tratamiento de aguas residuales y a fomentar su reautilización. La depuración eficaz de las aguas residuales se ve influenciada por la automatizaci ón en la captura de datos, ya que supone un aumento considerable de la información suministrada por los sensores para realizar los controles de las plantas. Disponer de toda la información, ordenarla y relacionarla de la manera más eficiente, mediante el análisis de datos, y obtener el conocimiento necesario para una correcta toma de decisiones, sirve para optimizar el proceso, predecir momentos críticos en la depuración y ajustar los costes de explotación y operación. El 70% de los costes totales de una EDAR se deben a la aireación en el proceso biológico por lo que resulta de vital importancia ajustar este proceso al máximo para conseguir que las aguas depuradas cumplan la legislación con el menor gasto posible. Este trabajo analiza los datos proporcionados por el benchmark BSM1_LT durante un periodo de un año y medio para realizar una clasificación. En la primera parte se realiza una clasificación de los datos en función del tiempo atmosférico debido a que la cantidad de contaminantes se ve influenciada por volumen de agua a tratar y las características ambientales y esto afecta al volumen de oxígeno que ha de ser aportado por las soplantes para conseguir que el amonio final se encuentre dentro de los límites exigidos por la normativas. Por otro lado se clasifican los valores de salida del nitrógeno orgánico o amoniacal, con ventanas temporales, para comprobar si el oxígeno suministrado es el adecuado o se puede modificar para optimizar los costes. Para mejorar la operatividad de la planta se realiza una predicción del valor del amonio del efluente, tambien utilizando ventanas temporales, lo que permitirá variar la cantidad de oxígeno proporcionado por las soplantes del sistema de aireación y ajustarlo a las necesidades del momento de manera eficiente.
In 2015, UN adopted the 2030 Agenda for Sustainable Development and its Sustainable Development Goals will mobilize efforts to finish all forms of poverty, tackle climate change, while ensuring that no one is left behind. Goal 6 ensure access to water and sanitation for everyone. World Water Day 2017 theme was wastewater and the reduce and reuse of wastewater. The efective purification of wastewater is influenced by the automation in data capturing, because it supposes a considerable increase of the information provided by the sensors to carry out the controls of the plants. Once you have all the information, order and relate it in the most eficient way, through data analysis, and obtain the necessary knowledge for a correct decision making, serves to optimize the process, predict critical moments in the debugging and adjust the costs of operation. 70% of the total costs of an WWTP are due to the aeration in the biological process, so it is vital to adjust this process the most you can, in order to achieve that the treated water complies with the legislation with the lowest cost. This paper analyzes the data provided by benchmark BSM1_LT during a period of one and a half years to make a classification. On one hand, a classification of the data according to the weather is carried out because of the quantity of pollutants is influenced by the volume of water to be treated and the environmental characteristics. This affects to the volume of oxygen that has to be contributed by the blowers to ensure that the final ammonium is within the limits required by regulations. On the other hand, the output values of organic or ammoniacal nitrogen are classified using temporary windows, to check whether the oxygen supplied is adequate or can be modified to optimize costs. To improve the operability of the plant, a prediction of the ammonium value of the efluent is made, with temporary windows too, which will allow to vary the amount of oxygen provided by the aeration system blowers and adjust it eficiently.
In 2015, UN adopted the 2030 Agenda for Sustainable Development and its Sustainable Development Goals will mobilize efforts to finish all forms of poverty, tackle climate change, while ensuring that no one is left behind. Goal 6 ensure access to water and sanitation for everyone. World Water Day 2017 theme was wastewater and the reduce and reuse of wastewater. The efective purification of wastewater is influenced by the automation in data capturing, because it supposes a considerable increase of the information provided by the sensors to carry out the controls of the plants. Once you have all the information, order and relate it in the most eficient way, through data analysis, and obtain the necessary knowledge for a correct decision making, serves to optimize the process, predict critical moments in the debugging and adjust the costs of operation. 70% of the total costs of an WWTP are due to the aeration in the biological process, so it is vital to adjust this process the most you can, in order to achieve that the treated water complies with the legislation with the lowest cost. This paper analyzes the data provided by benchmark BSM1_LT during a period of one and a half years to make a classification. On one hand, a classification of the data according to the weather is carried out because of the quantity of pollutants is influenced by the volume of water to be treated and the environmental characteristics. This affects to the volume of oxygen that has to be contributed by the blowers to ensure that the final ammonium is within the limits required by regulations. On the other hand, the output values of organic or ammoniacal nitrogen are classified using temporary windows, to check whether the oxygen supplied is adequate or can be modified to optimize costs. To improve the operability of the plant, a prediction of the ammonium value of the efluent is made, with temporary windows too, which will allow to vary the amount of oxygen provided by the aeration system blowers and adjust it eficiently.
Descripción
Categorías UNESCO
Palabras clave
aprendizaje automático, clasicación supervisada, clasicación no Supervisada, ARIMA, depuración de aguas residuales, Nitratos, BSM1_LT, optimización energética, machine learning, supervised learning, clustering, wastewater treatment, Nitrate, energy efficiency
Citación
Centro
Facultades y escuelas::E.T.S. de Ingeniería Informática
Departamento
Inteligencia Artificial