High-speed fault detection and location in DC microgrids systems using Multi-Criterion System and neural network

摘要

This paper presents a new protection method for LVDC ring-bus microgrid systems based on Multi-Criterion System (MCS) and Neural Network (NN). The proposed method aimed at high-speed detecting line-to-ground (LG) and line-to-line (LL) low impedance faults without using a definite threshold of differential current by using specific rules and multi-criterion system. MCS protection showed speed and accuracy compared to differential protection. Also, NN estimated fault location in percent of line length acceptably as a secondary controller. In order to evaluate the reliability and the enforceability of fault detection and location schemes, simulated network and protection algorithms are implemented and tested in laboratory-scale. The implementation results indicate that the MCS and NN protection scheme can consistently detect and estimate fault locations in the order of a few milliseconds. To reach this goal, a loop type LVDC microgrid with proper power electronic equipment like solid-state bidirectional breakers and the multi-level inverter is fulfilled. (C) 2019 Elsevier B.V. All rights reserved.