Advanced signal processing algorithm for fault classification and localization in VSC-HVDC based Offshore wind farm

• The study present a scheme for fault detection and classification in different sections of offshore wind farm. • Real time execution of faulty section and its localization leads to accurate classification. • Signal processing methods used are Stockwell Transform and Ensemble Empirical Mode Decomposition • A quantitative calculation is included with classification accuracy and fault localization error. This work provides real-time validation on the RTDS platform by the use of S-transform, Ensemble Empirical Mode Decomposition (EEMD), and SVM, for quick and reliable detection of AC/DC faults. Next for classification, features extracted through intrinsic mode function decomposed by EMD and EEMD and classified distinctly using support vector machine techniques. The simulation results reveal that S-transform and IMF1-H in association with MPNN and LSSVM can effectively detect and classify AC/DC faults even under raw signal conditions. This paper also presents the fault localization in the high-voltage direct current cable line connected to OWF by traveling wave and EEMD. The detection and classification are carried out on an offshore wind farm (OWF) system integrated to an onshore grid through a voltage source converter-high voltage direct current (VSC-HVDC) in MATLAB, as well as in RTDS(real time digital simulation).