Title: Detection and classification of faults in DC microgrids utilising artificial neural network with bidirectional gated recurrent units

Authors: Ramaprasanna Dalai; Sarat Chandra Swain

Addresses: School of Electrical Engineering, KIIT University, Patia, Bhubaneswar, Odisha, 751024, India ' School of Electrical Engineering, KIIT University, Patia, Bhubaneswar, Odisha, 751024, India

Abstract: This paper proposes a fault detection and classification method for DC-MG. This paper aims to identify and categorise errors with increased precision. In order to improve the detection accuracy, a hybrid artificial neural network (ANN) and bidirectional gated recurrent unit (BiGRU) is proposed. The input signal's features (voltage and current) are extracted using the discrete wavelet transform (DWT). The hybrid ANN-BiGRU model analyses and classifies faults in the microgrid system using the features extracted using the DWT. The osprey optimisation algorithm (OOA) is used to adjust the weight parameters to minimise error. The simulation results were obtained using the MATLAB/Simulink tool. The simulation results indicated that the proposed technique acquired a higher accuracy of 99.65%, precision of 99.60%, and recall of 98.5% compared to other state-of-the-art methods.

Keywords: DC microgrid; DC-MG; osprey optimisation algorithm; OOA; artificial neural network bidirectional gated recurrent unit; ANN-BiGRU; discrete wavelet transform; DWT; renewable energy sources; RESs.

DOI: 10.1504/IJPELEC.2025.143953

International Journal of Power Electronics, 2025 Vol.21 No.1, pp.1 - 27

Received: 10 May 2024
Accepted: 19 Oct 2024
Published online: 15 Jan 2025 *