Research On Submarine Pipeline Tracking Technology Based On AUV Equipped With Forward Looking Sonar

The development,storage and transportation technology of China’s offshore oil and gas resources has developed rapidly.The increasing number of submarine pipelines has put forward higher requirements on their inspection methods.At present,pipeline inspection mainly relies on survey vessels and remotely operated vehicles(ROVs)to conduct section-by-section inspection along the pipeline,which is a huge workload and extremely costly,and requires intelligent solutions.With the development of science and technology,Autonomous Underwater Vehicles(AUVs),which can replace humans in underwater operations,have emerged and gradually become the main tool in the field of marine exploration.At the same time,the application of underwater acoustic detection technology is also very extensive and effective,especially the emergence of forward looking sonar.Therefore,it is a very effective way to carry out real-time pipeline inspection by AUV with forward looking sonar.In this study,based on the summary of related research at home and abroad,the submarine pipeline tracking technology is studied,and the pipeline tracking is realized in real time by using AUV with forward looking sonar without the assistance of mother ship.The main contents of this research are as follows.(1)The two-dimensional pseudo color sonar image of the submarine pipeline is obtained by using AUV with forward looking sonar,and converted into 8-channel grayscale image,then the segmentation and binarization algorithm of the sonar image is used to obtain the binarized image of the submarine pipeline,and the pre-processing of the sonar image of the submarine pipeline is realized.(2)To address the problem of large computational volume of pipeline image refinement processing,an optimized Rosenfeld thinning algorithm is proposed,which effectively eliminates the redundant data in the binary image and shortens the thinning time of the pipeline image.The algorithm is based on the principle of the classical Rosenfeld algorithm,combined with the characteristics of the binary image of the submarine pipeline,and reduces the thinning range to contain only the effective pixel range of the pipeline image,and iterates the computation only in this range.(3)Based on the refined pipeline image,a method is proposed to obtain the coordinates positions of the target points to be tracked on the pipeline image in real time to achieve pipeline routing.The method uses full least squares to fit the thinned pipeline image,discriminates straight pipelines from curved ones by judging the distance sum of all pipeline pixel points to the fitted straight line,and obtains the pipeline routing by straight line fitting and power function fitting for straight and curved pipelines,respectively.(4)The submarine pipeline tracking trial was conducted by using AUV with forward looking sonar and based on the pipeline tracking algorithm developed in this paper.The results of the sea trial show that the proposed pipeline tracking technology can achieve accurate and stable submarine pipeline routing tracking.