Antonio Morán

Antonio Morán Álvarez

 

a.moran@unileon.es

I received the degree of Dr. from University of Oviedo in 2012 and the MEng degree from the University of León, Spain, in 2007, in Industrial Engineering. I have been working as a research scientist at the Department of Electrical Engineering, Systems and Automatic Control at University of León since 2005. My current research interests are the design of new systems for virtual labs, remote control, supervision of industrial process and energy efficiency.

Curriculum vitae

Thesis

Analysis and forecasting of energy consumption profiles in public buildings using data mining techniques.

Sustainable and responsible consumption of energy requires establishing measures which pursue higher energy efficiency. A heavy knowledge of the consumption patterns enables to establish the existence of unnecessary consumption, optimize the daily demand, balance the loads, etc. succeeding in reducing the consumption level. If the energy market liberalization is also taken into account, the extracted knowledge can be leveraged to improve the negotiation with the supplier company, thereby reducing costs. This thesis analyze a large energy infrastructure, such as the University of León, characterized by having several measuring points (30 buildings) of high consumption which are scattered, so the data analysis becomes a complex task.
Throughout this paper, it is proposed the development of a technological infrastructure to gather data related to energy consumption, store and process them and show them later to an end user. This architecture implements a three-tier structure, transparent and modular, which can be used with different types of energy consumption such as electrical, heating or cooling systems. In addition, data are collected from environmental variables which have an influence on the level of consumption. They will be useful in the phases of analysis, interpretation of the results and forecasting.
This infrastructure includes analysis tools to generate visualizations and models in order to extract implicit knowledge in the energy data. Using the most representative consumption variable, which is the active power in this thesis, profiles of the daily dynamic of each building are generated. The analysis of the profiles or curves of consumption by dimensionality reduction techniques such as the CCA, SNE and t-SNE allow 2D projections of the profiles to be obtained for visual analysis and data comparison. This way, it can be established which buildings have a similar behavior and which ones have deviations from normal behavior. The analysis of these curves and the environmental variables also determines which ones have a greater influence on consumption. It is possible to create models using these variables along with other indicators of the consumption level, such as an activity index that models the usage level of a building. These models can be created using techniques such as SOM, they represent the consumption levels and allow a comparison between them.
In addition, methods for short and medium term forecasting of electricity consumption have been proposed. The obtained models make it possible to estimate the consumption of a building in the medium term (1 month) using the relationship between environmental variables and consumption profiles. From the future estimation of environmental variables, it is possible to forecast the consumption of a building and establish thresholds for the consumption levels
In the case of short-term prediction techniques, dimensionality reduction techniques, or modifications thereof, are used together with prediction techniques. Since there are different possibilities, in this thesis, four techniques are tested (CCA + ARIMA, SOM MIMO, KSOM and Neural Gas + SVR) and compared with each other to determine which is best. The use of these techniques involves the use of regressors (inputs to the prediction model) that incorporate information on the evolution of consumption. To achieve greater accuracy is required to choose the best regressor using selection techniques, such as k-NN, which does not need to train the algorithm, in order to save time. Depending on the dynamics of the building and how these have been grouped, the regressor is modified, and the accuracy of the prediction method changes.
To complete the consumption study, the data from the supplier company bills are also analyzed, in order to check the behavior of the billing points regarding to buildings. This analysis provides information about the company penalizations to the University for high or different consumption to the contracted one. Furthermore, it allows to establish performance measures, such as power factor correction or contracted power.
Finally, a set of visualization tools has been developed with Processing, and they are accessible via the Internet, so that the methods of data analysis and forecasting which have been proposed can be understood by an unskilled person in data analysis. These tools are interactive so the data to visualize can be modified and knowledge extraction becomes an easier task.

Sponsored projects

• Name Supervisión de Procesos Industriales Utilizando Técnicas de Visualización Basadas en Modelos de la Dinámica. (Industrial Process Monitoring Using Visualization Techniques Based on Dynamics Models)
DPI2009-13398-C02-02.
• Institutions Universidad de León, Universidad de Oviedo
• Main Researcher Manuel Domínguez González
• Donor Ministerio de Educación y Ciencia. CICYT (Ministry of Education and Science)
• Name Desarrollo de interface HMI. (Devolopment of a HMI interface.) SEDUCE
• Institutions Universidad de León (OPI colaborador de Indra)
• Main Researcher Manuel Domínguez González
• Donor Centro para el Desarrollo Tecnológico Industrial
• Name Obtención de modelos dinámicos utilizando tecnicas de proyección para mejorar la eficiencia energética en edificios del sector terciario. (Obtaining dynamic models using projection techniques to improve energy efficiency in buildings of the tertiary sector.) ULE 2009-4
• Institutions Universidad de León
• Main Researcher Juan José Fuertes Martínez
• Donor Universidad de León (University of Leon)
• Name Red Temática e-Automática. (Thematic Net e-Automatica) DPI2006-27217-E
• Institutions Universidad de León, Universisdad de Almería, Universisdad de Alicante, Universidad Nacional de Educación a Distancia, Universidad Miguel Hernández, Universidad Politécnica de Valencia, Universidad Politécnica de Cataluña.
• Main Researcher Sebastián Dormido
• Donor Ministerio de Educación y Ciencia. CICYT (Ministry of Education and Science)
• Name Excavaciones Arqueológicas del Castro de los Judíos de Puente Castro.(Archaeological Excavations of the Jew Village in Puente Castro) Código AD-67
• Institutions Ayuntamiento de León, Junta de Castilla y León, Universidad de León
• Main Researcher José Luis Avelló Álvarez
• Donor Junta de Castilla y León (Autonomous Government of Castilla y León)

Research experience

• Dates 2008 - Present
• Position held Research Fellow
• Institutions Universidad de León. (University of Leon)
• Donor Junta de Castilla y León (Autonomous Government of Castilla y León), cofinanced by the European Social Fund.
• Objectives Knowledge extraction based on dimensionality reduction techniques for analysis and forecasting of energy consumption profiles.
• Dates 2007 - 2008
• Position held Research Fellow
• Institutions Instituto de Automática y Fabricación (Automation and Fabrication Institute)
• Donor Instituto de Automática y Fabricación (Automation and Fabrication Institute)
• Objectives Knowledge extraction techniques based on visual data mining for the remote supervision of industrial processes.
• Dates 2006 - 2007
• Position held Colaboration Fellow
• Institutions Universidad de León. (University of Leon)
• Donor Minsterio de Educación y Ciencia. (Ministry of Education and Science)
• Objectives Modeling and simulation of an assembly line for practices with an industrial robot and pneumatic systems.
• Dates 2005 - 2006
• Position held Colaboration Fellow
• Institutions Instituto de Automática y Fabricación (Automation and Fabrication Institute)
• Donor Universidad de León. (University of Leon)
• Objectives Programming of control algorithms and PLC configuration for use in practices of industrial applications.

Publications

Articles and book chapters

Time Series for fault detection in an industrial pilot plant. A. Morán, J.J. Fuertes, S. Alonso, C. Masa, M. Domínguez. International Journal of Modern Physics B 26(25) 1246003(1-6). World Scientific Publishing Company, 2012. ISSN: 0217-9792. Ranking (2012): 118/125. JCR.
DOI: http://dx.doi.org/10.1142/s0217979212460034
Monitoring power consumption using a generalized variant of Self-Organizing Map. S. Alonso, A. Morán, M.A. Prada, P. Barrientos y M. Domínguez. International Journal of Modern Physics B 26(25) 1246005(1-9). World Scientific Publishing Company, 2012. ISSN: 0217-9792. Ranking (2012): 118/125. JCR. DOI: http://dx.doi.org/10.1142/s0217979212460058
Two-stage approach for electricity consumption forecasting in public buildings. A. Morán, M.A. Prada, S. Alonso, P. Barrientos, J.J. Fuertes, M. Domínguez, I. Díaz. Advances in Intelligent Data Analysis XI. Lecture Notes in Computer Science, 2012, Volume 7619/2012, ISSN: 219-229 Springer Berlin/Heidelberg. Índice de Impacto: 0.034 (SJR)
DOI: http://dx.doi.org/10.1007/978-3-642-34156-4_21.
Analysis of electricity consumption profiles by means of dimensionality reduction techniques. Engineering Applications of Neural Networks. A. Morán, J.J. Fuertes, M.A. Prada, S. Alonso, P. Barrientos y I. Díaz. Communications in Computer and Information Science, 2012, Volume 311, 152-162. Springer Berlin/Heidelberg ISSN: 1865-0929. Índice de Impacto: 0.027 (SJR)
DOI: http://dx.doi.org/10.1007/978-3-642-32909-8
Monitoring industrial processes with SOM-based dissimilarity maps. Expert Systems with Applications. M. Domínguez, J.J. Fuertes, I. Díaz, M.A. Prada, S. Alonso y A. Morán. Elsevier 39 (2012) 7110–7120, 2012. ISSN: 0957-4174 Ranking(2011): 5/77 en JCR.
DOI: http://dx.doi.org/10.1016/j.eswa.2012.01.033
Remote laboratory of a quadruple tank process for learning in control engineering using different industrial controllers. M. Domínguez, J.J. Fuertes, M.A. Prada, S. Alonso y A. Morán. Computer Applications in Engineering Education, 2011. ISSN (ISSN)1099-0542 Ranking(2011): 71/90 JCR. DOI: http://dx.doi.org/10.1002/cae.20562
Visualization maps base on SOM to analyze MIMO systems. J.J Fuertes, M Domínguez, I. Díaz, M.A. Prada, A. Morán y S. Alonso. Neural computing and applications. Springer 22/06/2012. Wiley ISSN 0941-0643. Ranking(2011): 80/111 JCR.
DOI: http://dx.doi.org/10.1007/s00521-012-1090-3.
Application of SOM-based Visualization Maps for Time-Response Analysis of Industrial Processes.M.A. Prada, M. Domínguez, I. Díaz, J.J. Fuertes, P. Reguera, A. Morán y S. Alonso. Artificial Neural Networks – ICANN 2010. Lecture Notes in Computer Science, 2010, Volume 6353/2010, 392-401. Springer Berlin/Heidelberg ISSN: 0302-9743. Índice de Impacto: 0.034 (SJR)
DOI: http://dx.doi.org/10.1007/978-3-642-15822-3_47
Desarrollo de un Laboratorio Remoto para la Formación Vía Internet en Automática. ISBN 978-84-936618-3-0. 102 págs. 2009.

Conference papers

Visual continuous maps for electric bills comparison. CONTROLO’12
Virtual and remote laboratory of a DC motor. 9th ACE
Improving user interaction in remote laboratories through HTML5/AJAX. 9th ACE
Advanced tools for electricity consumption monitoring. ICREPQ’12
Time series for fault detection in an industrial pilot plant. 5th PHYSCON
Dimensionality reduction for damage detection in engineering structures. 5th PHYSCON
A variant of self-organizing map (SOM) for monitoring power consumption. 5th PHYSCON
Analysis of Electric Power Consumption Using Self-Organizing Maps. 8th IFAC
Remote Laboratory for learning of AC drive control. 8th IFAC
Lighting Management System using KNX in the Remote Laboratory of Automatic Control at the University of León. 4th KNX Scientific Conference
Simulación en Easy Java de Prácticas de Control Realizadas con Equipo Feedback MS-150. 6th EIWISA
Simulaciones Interactivas en Easy Java de Equipo Feedback MS-150. JA2010
Gestión vía Internet de la Iluminación en el Laboratorio Remoto de Automática de la Universidad de León. JA2010
Practical Simulation Environment using EJS in the Remote Laboratory of Automatic Control of the University of León. CONTROLO’10
Development of an educational tool in LabVIEW and its integration in remote laboratory of automatic control. 8th ACE
Tool for instruction and learning on variable speed drives through the Internet. ECC09
Inter-university network of remote laboratories. 17th IFAC

Other

Other certification

• Foxboro Control Software 3.0 Configuration Essentials. Invensys España, 2012.
• Expert en sistemas iRio. Schneider Electric España, 2011.
• LAbVIEW Associate Developer. National Instruments España, 2010.
• KNX Partner Associated. Schneider Electric España, 2009.
• Inspector Técnico de Construcciones. Colegio de Ingenieros Industriales de Asturias y León, 2009.
• Curso de Aptitud Pedagógica. Universidad de León, 2008.