Harmonic Source Identification in Microgrid using TQWT and SVM Classifier

The detection and characterization of harmonic sources within Microgrid is an essential responsibility of power engineers, given its direct impact on the system's stability and efficiency. Wavelet analysis has emerged as a highly effective method for detecting harmonic sources in power systems due to its ability to analyze signals across multiple scales. Utilizing a support vector machine (SVM) classifier and a discrete wavelet transform based on the Tunable-Q wavelet transform (TQWT), this article presents an innovative method for identifying harmonic sources in Microgrid. The proposed method extracts feature from the voltage signals of Microgrid using TQWT, which offers improved spectral resolution and time-frequency localization compared to other traditional wavelet transforms. The extracted features are then fed into the SVM classifier, which accurately classifies the sources of harmonic disturbances. The performance of the proposed method is evaluated on a simulated Microgrid model, and the results show that it achieves high accuracy in identifying harmonic sources. The proposed method can be a useful tool for Microgrid operators to identify and mitigate harmonic disturbances, thereby improving the power quality and reliability of Microgrid.