Projects / Programmes source: ARIS

Development and implementation of a method for on-line modelling and forecasting of air pollution

Research activity

Code Science Field Subfield
2.06.01  Engineering sciences and technologies  Systems and cybernetics  Control systems technology 

Code Science Field
T125  Technological sciences  Automation, robotics, control engineering 

Code Science Field
2.02  Engineering and Technology  Electrical engineering, Electronic engineering, Information engineering 
Dynamic systems, mathematical modelling, air pollution, Gaussian process models, artificial neural networks, air-pollution prediction.
Evaluation (rules)
source: COBISS
Researchers (7)
no. Code Name and surname Research area Role Period No. of publicationsNo. of publications
1.  11773  PhD Marija Zlata Božnar  Physics  Researcher  2013 - 2016  271 
2.  22483  PhD Dejan Gradišar  Systems and cybernetics  Researcher  2015 - 2016  161 
3.  27664  PhD Boštjan Grašič  Physics  Researcher  2013 - 2016  175 
4.  05807  PhD Nadja Hvala  Systems and cybernetics  Researcher  2014 - 2016  208 
5.  10598  PhD Juš Kocijan  Systems and cybernetics  Head  2013 - 2016  450 
6.  04290  PhD Primož Mlakar  Physics  Researcher  2013 - 2016  268 
7.  32444  Dejan Petelin  Systems and cybernetics  Technical associate  2013 - 2014  36 
Organisations (2)
no. Code Research organisation City Registration number No. of publicationsNo. of publications
1.  0106  Jožef Stefan Institute  Ljubljana  5051606000  90,590 
2.  2574  MEIS environmental cosulting d.o.o.  Šmarje - Sap  2271478  304 
Air pollution is a serious environmental problem in the world, as well as in Slovenia. To monitor the current condition of pollution, data from measuring stations and model spatial calculations are used; to predict pollution in the future, only the models which can be divided into still unreliable three-dimensional pollution representation models, and raster location pollution prediction models, can be used. The purpose of the research project is to develop a Gaussian-process modelling method and model for accurate ozone predictions at the most heavily burdened locations in Slovenia. To this end, we will combine the scientific experience of two groups that have extensive references in the development of such models. The main approach consists of real-time learning of a temporally variable model. For this purpose, Gaussian-process models (GPMs) will be used. This method is appropriate for the identification of very complex nonlinear processes according to the black-box method, and has proven extremely efficient in the field of modelling of complex, nonlinear dynamic systems. The second approach deals with models based on a multilayer perceptron artificial neural network (MLP) which is proven to be a universal approximator for a nonlinear system of functions of several independent variables. The MLP application methodology for the field of air pollution prediction has already been developed. For the GPM, we will redefine and upgrade the methodologies developed by the MEIS team for the purposes of MPL. The implementation of a new GPM development methodology in the field of atmosphere processes will represent the main scientific result. This method enables dynamic adjustment to the process, and has so far never been used in this field, except in a preliminary study performed by the Jožef Stefan Institute (IJS) which gave very good results. The applied project result will be the test environment - test bed - used for the elaboration and testing of prediction models. For real-time application, an efficient ozone-concentration prediction system for selected locations across Slovenia will be elaborated. The developed method and the resulting algorithm will be evaluated mainly based on the measurement data from the coastal ozone measurement stations, where pollution is the most problematic. New efficient models developed within the scope of the project will be used for on-time and efficient alerting, which will result in better healthcare prevention measures and compliance to EU directives. The IJS and MEIS project consortium combines knowledge in the field of Gaussian-process modelling, experience in the field of air-pollution modelling, neural networks, and extensive experience in the field of environmental measurements. It possesses the required computer equipment, and the measurement data is publicly accessible through the Slovenian Environment Agency; and MEIS regularly produces its own detailed weather forecast.
Significance for science
The importance of project results on model development for ozone-concentration predictions across Slovenia with the nonlinear dynamic-system identification method, namely of Gaussian process models and artificial neural networks, is manifold in the development of the science. The main contributions are as follows: - Improved database with air-quality and meteorological data for multiple locations - Testbed and methodology for the validation of empirical models for ozone-concentrations prediction - Systematic empirical study of influential variables for ozone concentration for multiple locations - Various types of models, e.g., conventional regression, on-line, integrated, based on Gaussian process models and artificial neural networks with different purposes like prediction for mobile stations or improvement of existing theoretical models resolution - Publicly available results via publications about the developed methods as well as web page with predictions of air-quality - Introduction of a new type of radial frequency diagram – called pollution flower – for the analysis of air-quality data. To elaborate, the results of this project contributed to the development and validation of on-line modelling based on Gaussian processes of the actual and threatening problem – ozone air pollution modelling and pertaining predictions. Through comparative analysis, we selected a method most suitable in this case. Evolving Gaussian process models proved to provide most informative results. This was the validation of the method on realistic case study as well as contribution to air-quality prediction. Moreover, the algorithm based on this method enables mobility of the station that utilises the algorithm. As this study addressed multiple locations in Slovenia it represents a significant upgrade of the previous study, developed in the past for the city of Nova Gorica only. On the other hand deficiencies, mainly lack of measurement of some air-quality variables prevented use for even more locations. We investigated in what way and to what extent the scientific concept itself was universal to various locations and causes, and to identify the required adjustments according to the individual characteristics of the examined location. From the scientific viewpoint, the whole model system of ozone in Slovenia represents a substantial breakthrough in terms of quality compared to the current situation especially from the viewpoint of resolution. The developed integrated model between theoretical model and neural networks overcome the resolution of the Slovenian Environment Agency that publishes a forecast that is comprised of five categories for the administratively divided Slovenian territory regardless of diversity of the mechanisms connected to the urban or rural environment, valleys and hills and mountains, and others. The result that was not planned but has a significant importance is a new type of radial frequency diagram – called sunflower or pollution flower – for the analysis of periodical data. The sunflower or pollution flower diagram enables a quick and comprehensive understanding of the information about diurnal cycle of periodic data. It enables in a graphical form, quick screening and long-term statistics of big data when searching for their diurnal features and finding the differences between the data for several locations which was important for this project. We estimate the contributions as a step forward in the research of nonlinear systems identification.
Significance for the country
The project results are significant for Slovenia directly and indirectly. Relevance of the project to the development in Slovenia can be summarised as follows. - New empirical models are developed and validated for prediction of ozone concentration for some of Slovene cities where air-quality and meteorological measurements are available. - These models can be used for conveying information to general population. - The developed methodology can be potentially used for the modelling and prediction of some other variables of air-quality. - Research and development results will increase the competitiveness of involved partners and visibility in general. Developed models can supplement and assist the Slovenian Environment Agency in upgrading the existing ozone-pollution prediction system also for micro-locations. We expect that research results of this project will become prominent in Slovenia among the general and expert public. Based on the information on ozone from publicly available web page of MEIS company that was improved during the project the general public gains the awareness and proper planning of everyday activities, especially the ozone-sensitive section of the population. The quality of conveyed pollution information and the degree of risk to the population and subsequent timely notification of the ozone-sensitive population groups concerning ozone risks may bring new quality of life. A channel for performing the outreach of the project’s results is also involvement of project participants in educational activities mainly at the University level. The results will be used for upgrading the lecturing content at university level courses where the project participants collaborate. In general the research results and spanned publications are expected to increase the visibility of Slovene academic and SME research in the world. The project results are also extremely important to MEIS company because it will improve competitiveness of this Slovene company: - Through the project, the ozone prediction method they had been used before was upgraded from scientifically appropriate and professional aspect, enabling for the regular dissemination of data on the expected concentrations to the public. - The model system, which was the main project result, became a brand new MEIS product that is prepared for further marketing. - Real-time results of the new prediction models pertaining to this project are used for the improvement (through the assimilation method) of spatial pollution predictions for control of pollution in Slovenia.
Most important scientific results Annual report 2013, 2014, 2015, final report
Most important socioeconomically and culturally relevant results Annual report 2013, 2014, 2015, final report
Views history