| Since the insulators are exposed in the external environment for a long time,the pollutants in the air will accumulate on the surface.The contamination on the surface of the insulator will absorb the moisture in the air,which will result in a significant drop in the electrical strength of the insulator.When the flash-over occurs,it can cause serious power outages and line outages.In recent years,with the development of digital image processing and computer machine vision technology and the continuous advancement of “strong smart grid”,online monitoring and fault diagnosis of high-voltage transmission lines have gradually turned from manual to intelligent.Based on the image processing technology,this paper focuses on the identification and detection algorithm of contamination levels of porcelain insulators in bad weather conditions.This paper mainly carries out the following parts:(1)By analyzing the research status,inadequacies and development trends of the insulator pre-process and contamination detection technology,the research plan of this paper was formulated.Firstly this paper introduces the overall architecture to realize the contamination level detection technology of the transmission line insulators and explains the key technology blocks involved in the system.In addition,the overall implementation flow of the image processing algorithm for insulator contamination detection is introduced.(2)An improved MSRCR image enhancement algorithm is proposed for complex field environment and bad weather conditions of insulators.Firstly,the color image is decomposed into three single channels,and the reflective components of the image are extracted according to the proportion of each color channel in the original image.Then,two boundary points are calculated by two-dimensional minimum error method,and the reflective image is processed by adaptive threshold piece-wise linear transformation.Finally,the enhanced color image is obtained by combining the three channels.The H component is segmented to obtain the insulator target.(3)Establishing an insulator contamination levels detection model.By extracting and screening the color features of insulators,the characteristic parameters with strong classification ability are obtained as input of the model,and the corresponding contamination level of insulators is output to realize the identification and detection of the contamination levels of porcelain insulators.(4)The performance of relevant algorithms involved in insulator contamination levels detection technology is analyzed and tested.Including image enhancement algorithm,image segmentation algorithm and insulator contamination levels detection model.And field operation tests of insulators in different field environments are carried out.The resultsshow that the accuracy of the proposed method is 92%,which verifies the feasibility and practicability of the algorithm.This paper contains 38 charts,10 tables and 74 references. |
