Infrared Thermography-based Identification of Thermal Faults of Substation Equipment
In view of poor determining ability and low identification accuracy of existing identification methods,a thermog-raphy-based thermal fault identification method for substation equipment is studied.Based on Planck blackbody radiation technology,infrared temperature measurement is carried out to obtain infrared thermal images,and the images collected on site are preprocessed.Using multi-scale feature fusion,the fault features are extracted and normalized by equalized scaling to obtain vector machine sample data.The spatial sample data are mapped to a high-dimensional space by setting the optimal classification function for nonlinear transformation to obtain the optimal classification.Using kernel functions for identification under constraint conditions,the danger level of faults can be determined and identification results can be obtained.The experimental results show that the abnormal heating of the casing in three-phase can be specified as an e-mergency fault and constitute a defect.The identification and judgment are correct.After 5 tests,the accuracy of fault i-dentification is maintained at above 80%.The results meet expectations,achieving the utility of accurately identifying thermal faults in substation equipment.
NETL
NSTL
万方数据
