Application of multi-agent and data mining techniques in condition assessment of transformers

• Published in: 2004 International Conference on Power System Technology, 2004. PowerCon 2004 Vol. 1; pp. 823 - 827 Vol.1
• Main Authors: Lizeng Wu, Yongli Zhu, Xueyu Li, Jinsha Yuan
• Format: Conference Proceeding
• Language: English
• Published: IEEE 2004
• Subjects: Computer science; Costs; Data mining; Data warehouses; Helium; Maintenance; Power system reliability; Substations; System testing; Transformers
• ISBN: 0780386108; 9780780386105
• DOI: 10.1109/ICPST.2004.1460107

The framework of the assessment system for transformer condition is proposed using data warehouse, multi-agent system and data mining technologies synthetically. It is used to support the condition-based maintenance of transformer, improve the equipment reliability and reduce the substation operation cost. The health state of the transformer depends on the design, present and historical operation conditions of equipment. Usually, the off-line testing results, operation history and fault records of equipment, and weather conditions, etc. have been stored in different systems and they are very difficult to be utilized synthetically. Therefore, a data warehouse has been set up to integrate all of these data. We used data mining techniques to get the regulation and trend of the health condition of transformer and then make a conclusion. In order to implement the system efficiently and effectively, we employed some ideas of open agent architecture (OAA) to compose a multi-agent-based framework of the proposed system.