Partial Discharge in Power Transformers: Approaches for Estimating Number of Sources

Identifying partial discharge (PD) sources in power transformers ensures electrical systems' reliability and safety. In an AEPD-based monitoring system for identification of the number of sources, expanding the acoustic emission sensor array incurs elevated expenses, logistical complexities, and spatial limitations. Addressing these concerns is crucial, emphasizing the need to identify multiple PD sources with fewer AE sensors. In the case of multiple source single channel systems, Blind Source Separation (BSS) is one of the techniques commonly used to separate individual PD source signals. In BSS, identifying several source signals plays a vital role in the performance of the BSS algorithm. The present work simulates multiple PD source signals in a single channel in a MATLAB environment. Different techniques are attempted to identify the number of PD sources in the single-channel system.