Vibration Fatigue Damage Identification Of Offshore Platform Pipeline Based On Vibration And Acoustic Emission

With the continuous development of China’s economy,oil,natural gas,and other energy as the lifeblood of industrial production,its demand is also increasing,but gradually depleted land oil and gas resources,people just have to step up the pace of energy development.Offshore oil and gas production platforms are an important facility for exploiting and initially treating oil and gas resources.A large number of process piping are distributed on the platform to transfer the extracted oil and gas resources between equipment,platform,and land,which are the lifeline of the offshore oil and gas industry.However,the traditional process piping design is built on static analysis,and the vibration of piping is seldom considered.Long-term vibration will cause the risk of fatigue failure of piping,so it is very necessary to conduct damage detection for offshore platform piping.For the structure of offshore platform oil and gas piping system in the real environment,the piping system structure is huge and the piping direction is complex,the working environment is multifaceted,and damage detection is usually performed on the working state.Therefore,this paper researched the vibration-induced fatigue damage identification of process piping,based on vibration and acoustic emission damage identification as two branches,and based on this research,carried out damage identification research of multimode fusion to improve the accuracy of crack positioning.The main contents are as follows:(1)Research on the damage identification method of offshore platform piping based on vibration modal parameters combined with BP(back-propagation)neural network was conducted.Experiments were carried out on the structural model of a real-scale piping system as the research object.A large number of damage condition samples were obtained through complementation of experiment and simulation to verify the damage identification method.(2)According to the BP neural network in training when positioning error caused by the uncertainty of initial parameters,the improved BP neural network based on the genetic algorithm,and use of damage condition of the same sample to the original BP neural network and genetic algorithm to improve the BP neural network were trained and the damage localization,compare two methods of the training effect and position precision fitting.(3)The damage identification method of offshore platform piping based on the acoustic emission method is studied,and the damage identification is divided into signal feature identification and time-difference location.The acoustic emission test of offshore platform piping noise and a real-scale piping system structure was undertaken.On this basis,signal classification research was carried out,and the damage identification method based on EMD(empirical mode decomposition)combined with a PNN(probabilistic neural network)was verified.At the same time,acoustic emission fatigue crack location was carried out based on the test data,and the fatigue crack location method based on BP neural network modified timedifference method was applied.(4)Aiming at the accuracy of fatigue crack location,a multi-mode fusion damage identification method was proposed based on vibration and acoustic emission damage identification methods,and the effect of network location was verified by using test data.The training fitting effect and location accuracy of the BP neural network trained with single information were compared.