Spectral Line Analysis And Recognition Of Insulator Pollution State Based On Hyperspectral Technology

In recent years,with the rapid development of China's industry,environmental pollution is becoming more and more serious.Insulators in power system operate in heavily polluted environment,and are prone to flashover accidents in wet weather such as drizzle and so on.The safe operation of power system is facing a major threat.Hyperspectral technology is based on the physical and chemical properties of material,and is widely used in various fields,such as agriculture and forestry,and provides a new method for the recognition of material.The research of pollution state has important guiding significance for preventing pollution and cleaning dirty.Based on hyperspectral technology,this paper studied the relationship between spectral reflectance and pollution state parameters,and realized non-contact remote sensing monitoring of insulator contamination.A hyperspectral data analysis system was built to analyze the relationship between spectral characteristics and pollution state parameters of contaminated insulation sheets.Contaminated insulation sheet spectra showed substrate had no effect on the spectral characteristics of contamination.The spectral characteristics of contamination with different moisture levels were different.The reason was that the water itself has corresponding spectral characteristics,and the humidity can affect the spectral characteristics of contamination.The main reason that thickness affects contamination spectral information was that light can reflect through the thin contamination layer to the surface of the substrate.The test conclusion was applicable to the detection of contamination thickness of natural polluted insulators.On the basis of explaining the principle of different recognition methods,the methods of parallel hexahedron classification,minimum distance classification,martensitic distance classification,support vector machine and spectral angle mapping were used to identify the contamination of different components,kaolin on the surface of insulators,the existence of natural pollution and the degree of pollution.Comparing the accuracy of each recognition result,such as whether the contour of the polluted area is clear,and the possibility of false recognition,such as whether the background will be misidentified,the needs of the insulator contamination recognition in the project were consider,and the optimal identification method was selected.The spectral information of contaminant in different positions of the contaminated insulators was collected,and the relationship between the spectral reflectance and the actual pollution situation was analyzed.The spectral reflectance of different locations was compared,and the corresponding spectral information distribution characteristics were obtained.According to the corresponding relationship between the results of spectral reflectance comparison and the actual pollution degree,the characteristic bands of the pollution degree analysis were selected and the reflection characteristics of the insulating materials were used to explain.