Application Of Membrane Computing In Infrared Image Segmentation And Fault Detection For Electric Equipments In Transformer Substation

At the moment,the infrared thermal imaging technology which gains more and more people’s attention and favor,has been broadly used in on-line detection and fault detection of transformer substation power equipment.The infrared thermal imaging technology uses optical technology to detect heat radiation of infrared objects specific band signal,and then transforms the specific band signal into images which can be identified.For the intelligent power grid,in order to achieve on-line detection and fault detection,the infrared thermal imaging technology and the infrared image segmentation technology are necessary.Combining membrane computing,the infrared image segmentation technology and fault auto-detection of transformer substation power equipment has been made a deep study in this paper.First of all,the infrared image noise characteristics of substation is taken as a starting point,several existing image preprocessing methods are summarized.For different types of noise,the corresponding denoising method is selected to achieve the best effect of image preprocessing;Secondly,three kinds of traditional image segmentation methods and two kinds of clustering methods are discussed,the advantages and disadvantages of these methods are verified by concrete experiments,which have founded a good base for presenting a new image segmentation algorithm in the next step;Thirdly,several image segmentation algorithms of FCM are studied carefully,the advantages and disadvantages of these algorithms are summarized.Integrating the concept of membrane computing,a new membrane clustering algorithm for image segmentation based on FCM is presented(which is called P_FCM for short).Comparing to other algorithms,the superiority and effectiveness of P_FCM algorithm has been demonstrated by three different experiments;At last,the presented P_FCM is applied to infrared image segmentation and fault auto-detection of transformer substation power equipment,satisfactory results are obtained.The effectiveness and accuracy of the proposed P_FCM algorithm is verified again.