Research On Intelligent Monitoring System For Transmission Lines Against External Damage

In recent years,with the rapid development of various aspects of social production and life,China’s electricity demand has further increased,and the State Grid has increased its construction and investment in power transmission in order to meet the electricity demand of various regions.High-voltage overhead lines are the main and most widely used power transmission methods currently adopted by the State Grid.Its coverage is getting wider and wider.No matter in towns or villages,overhead transmission lines can be seen everywhere.Because the stability of the transmission line plays a decisive role in the user’s power supply status,it is particularly important for the monitoring and protection of the tower and transmission line.At present,China’s power grid maintenance is transitioning from manual inspection to remote video surveillance,but the degree of automation of video surveillance is far from enough,and often requires the intervention of background staff.Moreover,some intelligent recognition algorithms implemented in an ideal environment have a much lower intrusion recognition rate in bad weather than in a normal environment.According to the research status of the transmission line video surveillance system,the paper mainly made the following research work on some problems in the current video surveillance universal adaptability:(1)From two aspects of traditional image enhancement and physical modelbased image restoration algorithms,more advanced image defogging methods are studied.Based on the image defogging algorithm based on atmospheric light scattering model,a fusion of contrast optimization and edge perception is proposed.Weighted image defogging algorithm.The algorithm first improved the atmospheric light estimation in the atmospheric scattering model,by separating the sky area to complete the atmospheric light estimation more accurately.Then,in terms of transmittance estimation,the initial transmittance is estimated by balancing the local area contrast and information loss,and then the transmittance is refined by weighted guided filtering.Finally,the color shift that is easy to appear after defogging is corrected.After the image defogging is completed,it is compared with other classic defogging algorithms through the information entropy,average gradient and variance three non-reference image quality evaluation indicators to verify the effectiveness of the algorithm.Then through the O-HAZE data set and the peak signal-to-noise ratio(PSNR),structural similarity(SSIM)and other full-reference image evaluation methods,the image dehazing effect was evaluated more objectively.The effect of defogging and detail retention at the depth change is better,and the color fidelity of the image has been greatly improved.(2)According to the requirements of the video monitoring of transmission line anti-breakage,the research is conducted from two aspects: moving target detection and crane intrusion detection.For the detection of moving objects,through the analysis and comparison of the inter-frame difference method,the optical flow method,and the background modeling method,a method based on the mixed background Gaussian model is used to extract the picture foreground,and the model is improved to achieve the pixel Gaussian distribution Self-adaptive quantity,complete the detection of intrusive moving targets faster under the condition of ensuring accuracy;for crane intrusion detection,it is divided into edge detection and hough line detection,through the Sobel operator,Canny edge detection and based on Convolutional neural network edge detection and other methods are used for analysis.The overall nested edge detection method is used,which greatly improves the system’s ability to adapt to complex environments.After the edge detection,the system completes the detection of the boom target through hough line detection,and then realizes whether the boom is out of bounds.The algorithm also verifies the video and image after defogging,and can still accurately complete the corresponding intrusion detection.(3)Design and implement the condition monitoring system of the front-end line tower.In terms of hardware,the overall hardware circuit of the system is constructed,and the sensor selection is completed.The SCA100 T is used to detect the inclination of the tower.Its accuracy can reach 0.035°,and it has better longterm stability.The software has completed the construction of the μC/OSⅢ realtime operating system and the graphical interface based on em Win,realized the data collection of the tilt of the tower,and uploaded the data to the cloud server through wireless transmission,and realized the remote real-time monitoring of the status of the tower.