Research on condition assessment method of intelligent power transformer

• Published in: Proceedings of the 2011 International Conference on Electrical Engineering and Informatics pp. 1 - 4
• Main Authors: Feng-Jiao Wu, Guan-Jun Zhang, Shi-Qiang Wang, Hao Xu, Da Wang, Min Lei
• Format: Conference Proceeding
• Language: English
• Published: IEEE 01.07.2011
• Subjects: condition assessment; Indexes; membership function; Monitoring; Oil insulation; parameter weight; Power transformer insulation; Transformer cores; unascertained rational number
• ISBN: 1457707535; 9781457707537
• ISSN: 2155-6822
• DOI: 10.1109/ICEEI.2011.6021814

With the development of sensor and measurement technology, more and more parameters are monitored to assess power transformer condition. Different parameters may reflect different aspects of transformer insulation and the same parameter diagnosed by different methods may get different results. Facing such a complex transformer system and so much online and offline data, the most important thing is to find an effective method to fuse all important information to evaluate transformer condition. In this paper, a comprehensive assessment model based on hierarchical fuzzy theory is proposed to evaluate transformer condition. In the model, a transformer is divided into different components based on its structure such as winding, capacitance bushing, iron core, insulation oil, cooler, tap-changer and relay devices, and these parts usually meet some faults during operation. The transformer condition is assessed by its components while their conditions are evaluated by related monitoring parameters. While employing the model to assess the condition of a transformer, the key is to determine the weights and membership functions of different parameters. An unascertained rational number method is introduced to determine the different weights and some effective methods are proposed to determine the membership functions according to experts' experience and guidelines. Finally, an example is give to verify the effectiveness of the model.