Detection of Sensor Fault in a DC Microgrid Using Supertwisting Observer

Abstract

AbstractThe detection of sensor faults in a direct current (DC) microgrid is essential to provide a safe and uninterrupted supply of power. The fault detection techniques in the DC microgrid should involve immediate sensor fault detections to protect the system from adverse effects. To address this issue, in this paper, a fault detection technique is designed which addresses the scenario where the interconnected distributed generation units (DGUs) are affected by sensor faults in a DC microgrid. Multiplicative faults in the grid-connected voltage and current sensors are considered. A model-based multiple sliding mode observers (HOSM) are designed to estimate the states of the DGUs, and faulty sensors are detected based on the generated residuals. The stability of the overall microgrid system can be achieved by designing a hierarchical controller which ensures uniform current sharing among all the DGUs. The validation of the proposed method is supported through extensive MATLAB simulations for various arrangements of the DC microgrid subjected to numerous known and unknown disturbances. A common threshold is selected from the multiple simulations for sensor fault detection.KeywordsSensor faultsFault detectionDC microgridDGUsHOSM observerLMIRenewable energy