The Study On Diagnosis Method Of Substation Equipment Thermal Faults Based On Small Sample

The state of substation is rocketing with the expanding of grid, and insuring the safe running of the equipment plays an important role to guarantee the safe and stable operation of power system. Thermal appearance of substation equipment is a sign of fault, and the proportion of thermal fault of substation equipment failure quantity is the largest. Hence, it is necessary to diagnose the thermal fault. Meanwhile, we have insufficient fault samples in the diagnosis process of substation equipment fault. Due to the longer inspection cycle of the Station equipment, people collected less data. Because of the rarity of the equipment failure accident, the data collection in the scene is also very difficult. All in all, it increased the difficulty of the fault diagnosis for substation equipment. Therefore, the author selected substation equipment thermal fault as the research object, based on the small sample data, to analyzing the substation equipment fault diagnosis methods.In this paper, the author first deeply analyzes the mechanism and law of development of substation equipment thermal fault. On the basis of FMEA technology, the authors establish a substation equipment system definition and thermal failure mode analysis in order to provide the theory basis for thermal fault diagnosis for substation equipment. Secondly, in order to solve the problem of small sample classification, the author established a thermal fault diagnosis model based on SVM. The author using the improved particle swarm optimization algorithm to optimize a diagnosis model and used the actual data to verify accuracy and practical value of the optimum model diagnosis. Thirdly, In order to find the possibility of early thermal failure of the equipment, the author also adopted the SVR and multivariate time series models to forecast the trend of the characteristics. The establishment of thermal transformer substation equipment fault diagnosis system is of great help of Instant online heat equipment fault diagnosis.