Research On Visible Image Detection Technology Of Composite Insulators Power Transmission Line

Composite insulators have been widely used in the power system, due to its superior electrical characteristics and they have played an important role.to ensure the safe operation of transmission lines. With increasing of the running time, the hydrophobicity degradation of the composite insulators will appear. and the mechanical strength, electrical performance degradation, aging and other issues will also appear, this may lead to the transmission line faults. To ensure the safe operation of transmission lines, inspection and testing should take to the composite insulator working in the serious environmental pollution region. Traditional detection method of composite insulators has low detection efficiency, and insulator defects may be neglected. In this thesis, the transmission line composite insulators detection technoLogy based on visible light image analysis and detection was introduced, and it can be an effective solution to the above problems. This insulator detection method is security to works, and it can effectively reduce the transmission line operation, maintenance staff strength, improve the detection efficiency.Through analysis and comparison of the image preprocessing technique, such as the gray image transformation, image enhancement, image denoising, we propose using the gray-scale image in the insulators detection of the hydrophobicity and cracks. The histogram equalization and adaptive filtering technique is applied for the insulators image preprocessing.With analysis on the edge detection technoLogy, such as the edge detection differential operator, the Log operator and Canny operator etc., the differential operator edge detection algorithm is presented in the thesis to achieve the composite insulators surface cracks image detection and crack length i image measurement.In this thesis, the two-dimensional entropy thresholding method for hydrophobic image segmentation and mathematical morphoLogy theory were applied with the hydrophobic image processing. With the application of the morphoLogical binary image reconstruction, the image feature of insulator hydrophobicity can be extracted. Through the IEC standards at all levels of hydrophobic image analysis, the composite insulators hydrophobic template which can indicate HC1～HC7hydrophobic level was established. And this makes the composite insulators hydrophobic image measurement possible.In this thesis the color image analysis techniques were directly used in the composite insulators aging and contamination detection. With the application of improved watershed algorithm for color image segmentation, the insulator contamination area can be extracted. The HSl, RGB, LUV color spaces are calculated to achieve insulator recency and contamination characteristics in the thesis.To determine connection between the image characteristics of composite insulators and their electrical characteristics, electrical characteristic test of the aging composite insulators were carried out. The test results show that the aging state of composite insulators have no apparent connection with their electrical characteristics.An online monitoring system of the composite insulators was designed in this thesis, The composite insulators image properties, such as the recency, dirt accumulation and were tested in the laboratory. The Laboratory test results show that the design of the online monitoring system can meet requirements of the insulator visible image detection.Through the studies in this thesis, the aim for image measurement of composite insulator features, such as cracks, hydrophobicity, contamination and aging, has achieved successfully. The study results of this thesis are important to the analysis and assessment of the composite insulators state.