541 / 2017-05-13 14:15:48

Vibration Monitoring of Converter Transformer On-Load Tap-Changers Using Phase Space Reconstruction and Poincare Section

14112,9712,12925,14111,14110

On-load tap-changer (OLTC) is the most indispensable part in converter transformers extensively used for the voltage regulation. During its operation, the OLTC switches from one top position to another top position with switching contacts opened and closed successively. Such kinds of frequent operations of OLTC would eventually lead to the damages to switching contacts and other components, and in turn pose a threat to the reliability of converter transformer. Hence, it is of great benefits for utilities to investigate effective techniques for the mechanical condition monitoring of OLTC in converter transformer.

Over the past decades, vibration analysis method is becoming a hot issue to monitor the mechanical condition of OLTC. By detecting the changes in vibration signal's magnitude and/or event time in OLTC operation, it is possible to detect the OLTC condition to a certain degree. Meanwhile, the methods of continuous wavelet transform (CWT), the combination of CWT and self-organizing map, orthogonal decomposition of the wavelet and envelope analysis based on Hilbert transform are applied to extract the main features of vibration signals for the detection of OLTC condition. However, the mechanical structure of OLTC in converter transformer is more complicated. It is still difficult to fully describe the abundant information contained in vibration signals which are highly non-linear and time-varying with the features of randomness and overlapping.

According to the existed studies, there are some chaotic features in the vibration signals of OLTC indeed, which is further verified with the calculation of the largest Lyapunov exponent. Based on the phase space reconstruction theory, this paper first describes the methods of determining the delay time and embedding dimension of the phase space for the measured the vibration signals of some real OLTC in converted transformer in field and is shown in Fig 1. Here, the C-C method and Cao method are selected to obtain the delay time the embedding dimension separately with the advantage of high accurateness and noise immunity. With the reconstructed phase space of the vibration of the OLTC, the key problem is to explore the features of high dimension curved surface accurately and quantitatively. In the paper, a Poincaré section is carefully constructed with the reconstruction phase space to seek for the distribution features of the phase point in high-dimension phase space. Generally, a Poincaré map can be interpreted as a discrete dynamical system with a state space that is one dimension smaller than the original continuous dynamical system. Because it preserves many properties of periodic and quasiperiodic orbits of the original system and has a lower-dimensional state space, it is often used for analyzing the original system in a simpler way. With the Poincaré section, the evaluating index is defined to distinguish the operation condition of OLTC in converter transformer. Finally, the vibration measurement of some real OLTC in converted transformer in field is described briefly. With the proposed method, the operation condition of OLTC is capable of recognizing and the results are agreed well with the on-site maintenance results.