PD Source Localization Algorithm by Using SCMD Technique for Power Cables Condition Monitoring

Precise localization of partial discharge (PD) events in power cables is crucial for maintaining the safety and reliability of medium-voltage (MV) electrical systems, especially amidst noise and signal complexities. This study highlights the pivotal role of PD localization in mitigating risks such as downtime, fires, and grid overload resulting from PD-related issues like surface tracking discharges and void discharges. This study introduces the segmented correlation minimum dispersion (SCMD) algorithm. This novel approach advances beyond the existing segmented correlation trimmed mean filtering (SCTM) approach by addressing its limitations in noise sensitivity and signal distortions. SCMD integrates discrete wavelet transform (DWT), segmented correlation (SC), and minimum dispersion (MD) techniques for effective PD signal denoising and robust source localization. In the SC phase, segmented correlation operations generate precise correlation indices to estimate initial PD locations. The MD phase enhances accuracy by evaluating dispersion across segments, filtering out erroneous reflections, and refining estimates using high-precision averaged data. This dual-phase structure allows SCMD to significantly improve localization reliability, even in low signal-to-noise ratio (SNR) environments. Rigorous assessments through MATLAB simulations demonstrate the SCMD algorithm’s outstanding accuracy, achieving minimal error rates of 0.13% even under reduced SNR, making SCMD a robust tool for effectively identifying and locating PD events in power cables and improving the reliability and safety standards of MV electrical systems.