Aristotle University Of Thessaloniki (AUTH)
Υποστήριξη ελέγχου λειτουργίας σταθμών παραγωγής ηλεκτρικής ενέργειας με χρήση μεθόδων τεχνητής νοημοσύνης
Abstract
dc:descriptionThe present thesis introduces a methodology for the design and development of a supporting system for control monitoring of Thermal Power Plants (TPP), having as final objective the improvement of its performance. The thesis proposes the combined application of three research areas: Knowledge Engineering, Knowledge Discovery in Databases (KDD) and Intelligent Agents. The basic principles that were adopted and their adaptation to the needs of industrial systems are analyzed thoroughly. For the methodical capture of the domain knowledge (exploitation of the personnel’s implicit knowledge) and the proper specifications writing, a KE approach, CommonKADS, was applied. It resulted, among others, to a rule base for identifying false sensor measurements. The rules were applied both for cleaning the historical data that was used for the Data Mining (DM) and for the on-line validation of the sensor measurements. For deriving models, the KDD procedure was applied to historical TPP operation data. These models were used for the replacement of erroneous sensors measurements and for proving indications for actions to the TPP personnel. Particular attention was paid in the data pre-processing, as it is significantly important for deriving the most appropriate models. Various attribute selection methods and classification algorithms were compared through testing and finally the best were selected based on statistical results and execution times. The derived models were embedded to software agents. The Multi-agent System was structured in three layers, which are described analytically. Examples of the DM models used, statistical results and screens of the graphical user interfaces are presented for each layer. The simulation led to positive results that support the opinion that the proposed framework can constitute a supporting system for control monitoring of a TPP and an advisory tool for its personnel. It is argued that the operation parameters can be regulated via proper indications, so as the TPP performance to be among the best 10% achieved in the past, with significant economical and environmental benefits.
Degree
thesis:*- Grantor dc:publisher
- Aristotle University Of Thessaloniki (AUTH)
- Year dc:date
- 2009
Author and committee
dc:creator, dc:contributor.*- Authors dc:creator
-
- Athanasopoulou, Christina
- Αθανασοπούλου, Χριστίνα
Subjects
dc:subject × 16- Σταθμός παραγωγής ηλεκτρικής ενέργειας
- Παρακολούθηση ελέγχου λειτουργίας
- Μηχανική γνώσης
- Εξόρυξη δεδομένων
- Ευφυής πράκτορας λογισμικού
- Σύστημα υποστήριξης αποφάσεων
- Power plants
- Control monitoring
- Knowledge engineering
- Data mining
- Intelligent agents
- Decision support systems
- Επιστήμες Μηχανικού και Τεχνολογία
- Επιστήμη Ηλεκτρολόγου Μηχανικού, Ηλεκτρονικού Μηχανικού, Μηχανικού Η/Υ
- Engineering and Technology
- Electrical Engineering, Electronic Engineering, Information Engineering
Rights
- Language dc:language
- gre
Identifiers
dc:identifier.*- Identifier
- 10.12681/eadd/23461