Arc Fault Detection and Location Technique Based on State Space Modeling for DC Microgrid

Abstract

Due to the increasing demand for sustainable energy, the DC microgrid with high penetration of power electronic devices is developing rapidly. The DC arc fault protection technique is essential to ensure the safety of DC microgrids. Although there are arc-fault detectors in the photovoltaic system, their misjudgment rate will increase when it is applied in the DC microgrid with a large number of power electronic devices. Furthermore, the cost will also increase when it is applied in the DC microgrid with many branches. In this paper, a series arc fault detection and location method based on state-space modeling analysis is proposed to realize arc fault detection and location in DC microgrids. Initially, the topology and operation principle of power converters in a DC microgrid is analyzed and the state-space model is developed. In this way, the equivalent circuit of a DC microgrid under normal operation and an arc fault can be established. Since the inductor current and the voltage of the output capacitor in a branch of a DC microgrid under normal operation and arc fault are different, they can be analyzed to determine the operating condition of the branches of a DC microgrid. Thus, the residual of inductor current and voltage of the output capacitor between theoretical analysis and practical measurement can be calculated to detect and locate the arc faults. Besides, the over-threshold values can be integrated to avoid the load transient influence on the arc fault detection. Hence, an arc fault detection and location technique based on state-space modeling and residual analysis is developed. The feasibility and high accuracy of this technique are verified by simulation and experiments in a typical DC microgrid composed of two branches.