Research On Underwater Multi-Source Navigation And Positioning Technology Based On Submarine Pipeline Detection

With the vigorous development of offshore oil resources,submarine oil and gas pipeline,as an efficient means of crude oil transportation,has become the main way of oil and gas transportation in the process of offshore oilfield development.Because the safety of submarine oil pipeline has a significant impact on the marine ecological environment,the regular inspection and evaluation after the completion of its laying has become an indispensable work content in the submarine oil and gas pipeline project.At present,it is effective and feasible to detect the submarine oil and gas pipeline by using the acoustic,visual and magnetic sensors carried by the autonomous underwater vehicle.However,the detection of the submarine oil and gas pipeline puts forward higher requirements for the underwater navigation technology,which requires the autonomous underwater vehicle(AUV)to have the ability of high-precision autonomous navigation and positioning with long endurance and long range,It is convenient to effectively locate the corrosion points,leakage points and buried points of oil and gas pipelines.Firstly,based on the design requirements of AUV detection in submarine pipeline,this paper gives the overall design scheme of AUV integrated navigation system,designs and implements an integrated prototype integrated navigation system for underwater detection of submarine pipeline detection robot.The navigation part of the robot adopts strapdown inertial navigation system(SINS)Doppler velocity log(DVL)and ultra short baseline acoustic positioning system(USBL).Secondly,the underwater acoustic environment is analyzed.In order to estimate the transmission path and distance of underwater acoustic sensor acoustic signal in seawater,this paper uses bellhop tool to model the underwater channel.The real sound velocity profile data is input to bellhop to obtain the channel transfer function concerned by the system.Based on this,a velocity/position measurement data simulator is designed and implemented to verify the DVL and performance.Then,the error analysis,theoretical derivation and mathematical model of SINS/DVL/USBL integrated navigation system are carried out respectively,the appropriate state quantity and measurement quantity are selected,the state equation and measurement equation of centralized Kalman filter are established,and the simulation experiment of the integrated navigation system in a specific state is carried out.Unscented Kalman filter(UKF)is used to smooth the estimation results,and a delay state unscented Kalman filter algorithm estimated by position delay measurement information is used.Finally,in order to reduce the positioning error caused by the acoustic outliers in the measured distance and direction information of USBL,an intelligent active outlier detection strategy based on one class support vector machine(one class SVM)is proposed.The weighting factor is set for the training set to realize the convergence and smoothing of local trajectory in the process of information fusion,and the robustness and applicability of the proposed method are verified by simulation experiments.