Fault detection of power converters in shipboard microgrids

Abstract

In shipboard microgrids (SMGs), the complexity of power equipment continues to increase as the level of electrification rises, especially for all-electric ships where power converters play a critical role in ensuring power quality. Currently, modern ships incorporate integrated energy storage systems with high storage capacities, and bi-directional converters are being employed to reduce the number of devices in the SMGs. To this end, a device's reliability is determined by its internal components, including capacitors and switches, where short circuits and open circuits faults originate and disrupt the quality of power flow. In this paper, the wavelet transform of faults on power converters is analyzed by using effective machine learning techniques based on K nearest neighbors in order for power converter fault detection effectively. The power converters of a practical ferry SMG are selected for computer simulations to ensure and demonstrate the performance of proposed method.