Application of wavelet transform for recognition of acoustic signals of various defects of power transformers

Background: This study discusses the use of wavelet transform for processing acoustic signals when diagnosing power transformers. A simulator with various defects was used to explore the application of continuous wavelet transform. The study also outlines an algorithm for obtaining signal standards to identify different types of defects. Aim. To develop a mathematical model for constructing reference acoustic control signals when processing data from a simulator with various defects. These signals will be used to identify and diagnose defects in real power transformers during their operation. Methods. The study introduces expressions for signal transformation, presents acoustic control signals, and explains the formation of reference signals. It also describes the process of recognizing the closest signals in form, corresponding to different types of defects. Results. The key results include expressions for signal transformation, acoustic control signals, and the formation of reference signals. The closest matching signals were identified, enabling accurate recognition of different defect types. Conclusion. The research proposes a method for processing acoustic control signals, allowing the reproduction of various insulation defects of power transformers and the recognition of such defects according to the proposed parameters of the wavelet transform.