Abstract
Partial discharge (PD) monitoring is widely used in rotating machines to evaluate the condition of stator winding insulation, but its practice on a large scale requires the development of intelligent systems that automatically process these measurement data. In this paper, it is proposed a methodology of automatic PD classification in hydro-generator stator windings using neural networks. The database is formed from online PD measurements in hydro-generators in a real setting. Noise filtering techniques are applied to these data. Then, based on the concept of image projection, novel features are extracted from the filtered samples. These features are used as inputs for training several neural networks. The best performance network, obtained using statistical procedures, presents a recognition rate of 98%.
Highlights
In electric power industry, it has been observed a growing interest towards predictive maintenance [1] for obtaining shorter equipment downtime and lesser costs
An important predictive maintenance technique for rotating machines is the partial discharge (PD) monitoring and analysis, which allows for the evaluation of stator insulation condition [2], [3]
Statistical properties extracted from wavelet decompositions of ultra-high frequency (UHF) signals are used as input features for training an artificial neural network
Summary
It has been observed a growing interest towards predictive maintenance (condition monitoring) [1] for obtaining shorter equipment downtime and lesser costs. Statistical properties extracted from wavelet decompositions of UHF signals are used as input features for training an artificial neural network In these references, as well as in most papers in literature, artificial PDs are generated in laboratory conditions to train and test proposed algorithms. Statistical tests are carried out to assess the performance and to pick the best neural network for classifying the PDs. The contributions of this work are the following: (i) novel input features, based on the concept of projection; (ii) new metric for evaluating recognition performance; (iii) to cover PD recognition in hydro-generators using real-world data, which is more complex than prior works in literature. Automatic recognition of complex patterns is difficult [16], and their presence is expected due to the occurrence of multiple simultaneous PDs sources (of different classes) in rotating machines, and due to the intense noise and interference from power systems themselves [2]
Sign-in/Register to view highlights & summary 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 callMore From: Journal of Microwaves, Optoelectronics and Electromagnetic Applications
Disclaimer: All third-party content on this website/platform is and will remain the property of their respective owners and is provided on "as is" basis without any warranties, express or implied. Use of third-party content does not indicate any affiliation, sponsorship with or endorsement by them. Any references to third-party content is to identify the corresponding services and shall be considered fair use under The CopyrightLaw.
