A multi-criteria computer package for power transformer fault detection and diagnosis

作者：Roncero Clemente Carlos; Roanes Lozano Eugenio^*

来源：Applied Mathematics and Computation, 2018, 319: 153-164.

DOI：10.1016/j.amc.2017.02.024

• Power transformer
• Transformer fault
• Predictive maintenance
• Computer algebra systems
• Knowledge based systems
• Verification

A package in Maple that helps users in power transformers fault detection and diagnosis has been developed. Transformers are required throughout modern interconnected power systems. Their range comprises from a few kVA to over a few hundred MVA, both in low voltage and in high voltage electrical network. As they are considered the key element in such systems, several maintenance methods have been reported in the literature: dissolved gas analysis (DGA) technique, short-circuit impedance (SCI) measurement, frequency response analysis (FRA) and power factor testing among others. All of them have as main goal to increase its useful life; normally reduced from aging process, stress conditions or electrical faults. Besides, they require special measurement devices and the experience of engineers, in order to make a proper diagnosis. This paper firstly determines the requirements of these tests to be applied and coordinate their input data and their output (diagnoses and recommendations). Afterwards, the package developed, that guides the users throughout the diagnosis processes, automatizes data processing and returns of the different tests (underlining if any contradiction between them arises) is summarized. The method is extensible/scalable by means of adding new techniques on this field of application.

• 出版日期2018-2-15