Abstract
In this paper fault detection and classification scheme using K nearest neighbor (KNN) has been presented for AC/DC transmission line with DFIG (doubly fed induction generator). The power system model consist of 1000 MW synchronous generator, 510 MW wind far, three phase transmission line of 500 kV and 30km length and HVDC line of 1000 MW. The model has been implemented with the corresponding parameters in MATLAB and simulated for different scenario. The raw current and voltage acquired for the sensors have been processed by performing discrete wavelet transform to the derive the discriminatory attributes with maximum disparity between the healthy and faulty cases. Following feature extraction, the required tasks of fault detection and classification scheme in A C and HVDC lines have been achieved using KNN. Further, the efficacy of the proposed KNN based scheme has been validated for varying fault scenarios pertaining to wide variation in fault resistance, inception angle of fault and fault location. For all tested cases, the proposed scheme is able to achieve a fault detection and classification accuracy of 1 00%.
Published Version
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have