Open Circuit Fault Diagnosis and Fault Classification in Multi-Level Inverter using Fuzzy Inference System

Abstract

Multi-level inverters (MLIs) have been successfully used to integrated the renewable energy sources (RES) into microgrids. However, the operation of MLI is affected when an open circuit fault (OCF) or a short circuit fault occurs. Among these kinds of faults, there is a high prevalence of open circuit faults in MLI. Any fault in MLI must be identified and classified as soon as possible to maintain the reliability of the power supply. This work is focused on developing a Fuzzy Inference System (FIS) for detecting and classifying the open circuit faults in Cascaded H-Bridge Multi-Level Inverter (CHMLI), thereby improving the fault diagnosis accuracy and efficiency. In CHMLI, the gate pulse is generated by pulse width modulation (PWM) technique. The Mamdani Fuzzy Logic Controller (FLC) identifies and categorizes the different OCFs. Fuzzy logic rules are designed for detecting and classifying open circuit faults simultaneously using the fundamental Discrete Fourier components of voltage and current. Several combinations of open circuit faults have been studied in different switches of the MLI, along with the effect of fault inception angle. Furthermore, the test results support the feasibility of the proposed fuzzy-based fault diagnosis and classification scheme in a practical context. A real-time simulation obtained with the help of FPGA-based OPAL-RT 4510 demonstrates the robustness and effectiveness of the designed topology. All types and fault locations are considered in multiple cases of switch failure.