Incipient Fault Detection and Condition Assessment in DFIGs Based on External Leakage Flux Sensing and Modified Multiscale Poincare Plots Analysis
Shouwang Zhao,Yu Chen,Feng Liang,Sichao Zhang,Nadeem Shahbaz,Shuang Wang,Yong Zhao,Wei Deng,Yonghong Cheng
DOI: https://doi.org/10.1088/1361-6501/ad076a
IF: 2.398
2024-01-01
Measurement Science and Technology
Abstract:Although doubly fed induction generators (DFIG) are widely used, difficulties in early fault detection and severity assessment for inter-turn short-circuit (ITSC) faults are highly prominent. In this manuscript, a novel incipient fault detection and state assessment method based on external leakage flux sensing and modified multiscale poincare plots (MMSPP) is proposed. An external leakage flux sensor is placed on the axial-end position of the generator to monitor the presence and evolution of ITSC faults. Multiscale poincare mapping is a novel nonlinear tool that is further developed and modified using the normal cumulative distribution function and multiscale computing methods to capture the behavior evolution and changes in the generator's external leakage flux signals. The healthy indicator is based on the analysis of elliptical orbit features extracted from MMSPP, established by support vector data description with parameter optimization. The effectiveness of the proposed method was implemented and verified under an experimental environment on a 100 kW DFIG platform to detect incipient inter-turn short circuits (mainly considering the two-turn ITSC) and evaluate the performance degradation (three- to eight-turn ITSC) with 0%, 50%, and 100% different load conditions. The experimental results showed the viability of the approach and fault indicator for incipient fault detection and condition assessment of the wind generator's low inter-turn insulation faults and for relative quantification of ITSC fault severity.