National Technical University of Athens (NTUA)
Πρόγνωση ποιότητας της ατμόσφαιρας στην ευρύτερη περιοχή Αθηνών με τη χρήση νευρωνικών δικτύων
Abstract
dc:descriptionThe aim of this Doctoral thesis is the use of artificial neural networks technology (ANNs) as forecast models for the air quality prediction in the greater Athens area. The term "air quality prediction" includes both the prediction of air pollution during the year and the prediction of human thermal comfort-discomfort, during the hot season of the year in the greater Athens area. For this purpose sixteen (16) Multi-Layer Perceptron ANNs have been created and trained according to the Back Propagation Algorithm. Twelve (12) of the created ANNs predict the air pollution in seven (7) different areas of Athens city for the pollutants CO, NO2, SO2 and O3 and in five (5) different areas of Athens city for the particulate matter (PM10). The prediction also concerns the persistence of the air pollution phenomenon, i.e. the number of the consecutive hours during the day, where the concentration of the pollutants was greater than the concentration limits as established under the European Framework Directive 96/62 / EC. In all cases, prognosis concerns the air pollution and the persistence of the phenomenon for the next three (3) days. The next four (4) ANNs forecast the population thermal comfort-discomfort, as well as the persistence of the phenomenon, i.e. the number of consecutive hours within 24 hours, where the thermal discomfort of the population due to high temperatures exceeds 50% (one in two people feel discomfort due to heat). The forecast covers eight (8) different areas of the greater Athens area during the warm season (May-September) of the year. The depth of the prediction is the following three (3) days. For the ANNs training data covering the time period 2001-2004 were used and the prediction was done for the year 2005. In the case of air pollution prediction as well as of the prediction of the human thermal comfort-discomfort the results seem to be quite satisfactory at a significant statistical level of p <0.01. Thus, it appears that the ANNs can be used as forecasting models for the air quality, both at a research-education level and an operational level, for the adoption of appropriate short and long term measures in order to protect public health in the greater Athens area.
Degree
thesis:*- Grantor dc:publisher
- National Technical University of Athens (NTUA)
- Year dc:date
- 2009
Author and committee
dc:creator, dc:contributor.*- Authors dc:creator
-
- Moustris, Konstantinos
- Μουστρής, Κωνσταντίνος
Subjects
dc:subject × 14- Πρόγνωση ατμοσφαιρικής ρύπανσης
- Πρόγνωση θερμικής άνεσης-δυσφορίας
- Τεχνητά νευρωνικά δίκτυα
- Αθήνα
- Ελλάδα
- Air pollution forecasting
- Thermal comfort-discomfort forecasting
- Artificial neural networks
- Athens
- Greece
- Επιστήμες Μηχανικού και Τεχνολογία
- Επιστήμη Χημικού Μηχανικού
- Engineering and Technology
- Chemical Engineering
Rights
- Language dc:language
- gre
Identifiers
dc:identifier.*- Identifier
- 10.12681/eadd/18219