A wavelet-based hybrid islanding detection system applied for distributed generators interconnected to AC microgrids

Abstract

The modern distribution system widely uses renewable energy sources (RESs) as a distributed generation that introduces some problems like unintentional islanding, protection concerns, and reverse power flow. Islanding or fault occurrences must be detected and treated to ensure system stability. The literature presents several islanding detection methods classified as passive (PDM), active (ADM), or hybrid detection methods (HDM). HDMs present advantages of both PDM and ADM methods and overcome their inherent drawbacks. This paper proposes a modification on the CWT for accomplishing its real-time implementation (RT-CWT), and the use of its nonstationary signal analysis for implementing a wavelet-based hybrid islanding detection scheme (WB-HIDS). The evaluation of some power quality-related indices: power grid voltage, frequency deviation, grid impedance, and power angle enables the detection of islanding condition or fault occurrence. The simultaneous use of various indices mitigates the non-detection zone related to the PDM. Also, the use of a transient detection scheme for triggering the signal injection implementation minimizes undesired disturbances introduced by ADM schemes. Experimental results obtained from an ac microgrid laboratory setup validate the proposed solution and demonstrate its effectiveness.