Answers are grounded in this paper and cite their sources.
Microgrid in electric vehicles (EV) require high reliability as any malfunctioning may result in loss of human live. To ensure robust protection for micro-grid, intelligent continuous monitoring and fault classification mechanism is required along with existing traditional protection scheme. Industrial internet of things (IIOT) makes low-cost continuous health monitoring data from EV micro-grid available. Manually processing of these bulk data for fault classification is cost and time consuming, and impractical. Neural network (NN) and other artificial intelligence classification systems require time-consuming feature engineering. In this digest, a Deep Neural Network (NN) framework, Convolution Neural Network (CNN) are applied for automatic and rapid processing of bulk health monitoring data for high precision micro-grid fault classification. The proposed classification method learns from data autonomously and extracts features by identifying anomalies in converter and inverter output voltage, and DC-link capacitor voltage amplitude, phase angle and harmonic components. This CNN based fault classification method shows better performance compared to conventional artificial intelligence based fault classification systems. The proposed classification system is verified under different operating conditions.
No takes yet. Share an insight, caveat, or question.
Haque et al. (2018) studied this question.