Abstract

White noise is a random signal mathematically analogous to an impulse function in time and flat spectrum in frequency. Naturally, the white noise may be used as an alternative test signal for determining the frequency response analysis (FRA) of transformers. Such an attempt not only enhances the power of FRA testing but also enable development of an online diagnostic test method. At present, the concept of using white noise is well-defined, but the crucial knowledge of applying the same on a real-time power transformer is currently not available or at least in the public domain. The study attempts to bridge this gap between theoretical contribution and practical requirement. The mathematical concept, experimental illustrations, and onsite / field validation altogether provides a complete insight into the identified problem. In this regard, an experimental study that investigates on using white noise as a test signal in determining the frequency response of a two-winding transformer is initiated. First, the possibilities of application of theoretical and mathematical concepts of using white noise as a test signal for measuring the frequency response of winding-core assembly is verified. Following this, the white noise signal applied to the circuit model and the corresponding response is recorded and the arbitrary choice of data samples measured from the circuit model is ensured using brute-force approach. In this procedure, the number of samples are manually increased until the magnitude and natural frequencies of resonant peaks matches atleast within 2 % accuracy with conventional FRA measurements. This is performed to ensure the repeatability of the method. Later, the time and frequency dependent statistical properties of the measured signals are computed and the respective transfer function is obtained. Subsequently, the transfer function obtained from the white noise method and the respective natural frequency (tabulated) are compared with the values measured using conventional FRA method. Once this is accomplished, these findings are experimentally validated on an actual high voltage transformer.

Full Text
Published version (Free)

Talk to us

Join us for a 30 min session where you can share your feedback and ask us any queries you have

Schedule a call