Fault diagnosis of wind turbine power converter using intrinsic mode functions with relative energy entropy

• Published in: Circuit world 2024-01, Vol.50 (1), p.90-107
• Main Authors: Manikandan, R, Raja Singh R
• Format: Article
• Language: English
• Subjects: Alternative energy sources; Circuit design; Empirical analysis; Energy conversion; Energy requirements; Energy sources; Entropy; Failure; Fault detection; Fault diagnosis; Fourier transforms; Induction motors; Insulated gate bipolar transistors; Inverters; Localization; Methods; Power converters; Renewable energy; Renewable resources; Robustness; Semiconductor devices; Shutdowns; Signal processing; System effectiveness; Time-frequency analysis; Turbines; Wavelet transforms; Wind power; Wind turbines
• ISSN: 0305-6120; 1758-602X; 0305-6120
• DOI: 10.1108/CW-09-2022-0241

Purpose The purpose of this paper is to prevent the destruction of other parts of a wind energy conversion system because of faults, the diagnosis of insulated-gate bipolar transistor (IGBT) faults has become an essential topic of study. Demand for sustainable energy sources has been prompted by rising environmental pollution and energy requirements. Renewable energy has been identified as a viable substitute for conventional fossil fuel energy generation. Because of its rapid installation time and adaptable expenditure for construction scale, wind energy has emerged as a great energy resource. Power converter failure is particularly significant for the reliable operation of wind power conversion systems because it not only has a high yearly fault rate but also a prolonged downtime. The power converters will continue to operate even after the failure, especially the open-circuit fault, endangering their other parts and impairing their functionality. Design/methodology/approach The most widely used signal processing methods for locating open-switch faults in power devices are the short-time Fourier transform and wavelet transform (WT) – based on time–frequency analysis. To increase their effectiveness, these methods necessitate the intensive use of computational resources. This study suggests a fault detection technique using empirical mode decomposition (EMD) that examines the phase currents from a power inverter. Furthermore, the intrinsic mode function’s relative energy entropy (REE) and simple logical operations are used to locate IGBT open switch failures. Findings The presented scheme successfully locates and detects 21 various classes of IGBT faults that could arise in a two-level three-phase voltage source inverter (VSI). To verify the efficacy of the proposed fault diagnosis (FD) scheme, the test is performed under various operating conditions of the power converter and induction motor load. The proposed method outperforms existing FD schemes in the literature in terms of fault coverage and robustness. Originality/value This study introduces an EMD–IMF–REE-based FD method for VSIs in wind turbine systems, which enhances the effectiveness and robustness of the FD method.