Research On Monitoring And Analysis Method Of Ice-induced Vibration Hazard On Offshore Platform

The offshore platform in cold regions are subjected to sea ice every winter,which causes serious ice-induced vibration problem.It is very important to study the ice-induced vibration hazards to ensure safe operation on ice-resistant offshore platforms and reduce the losses caused by sea ice hazards in cold regions.This paper focuses on the structural health monitoring of offshore platforms in cold regions,the methods of ice vibration response monitoring,damage identification and ice vibration risk prediction for offshore platforms in cold regions are studied.The main work is as follows:Firstly,aiming at the monitoring methods of ice vibration response for offshore platform in cold region.Basing on the characteristics of ice vibration response and modal confidence criterion(MAC)come up with an optimal sensor placement method,particle swarm optimization(PSO)based on dynamic inertial weight is used to solve the optimal layout problem.The method was used to optimize the sensor layout of a booster station platform in Bohai Sea,and get the best space layout position.Finally,the computational time,the number of iterations and the optimization ability of particle swarm optimization based on dynamic inertia weight algorithm are compared with the genetic algorithm for the same layout problem,the applicability of particle swarm optimization algorithm based on dynamic inertia weight is verified.Aiming at the problem that offshore platform in cold regions may exist hidden damage under long-term ice-induced vibration,Hilbert-Huang transform was used to identify structural damage.Firstly,the structure of ice-resistant jacket platform in Bohai Sea was established by finite element software ANSYS,the transient dynamic analysis was carried out under the loads of ice.Then Hilbert-Huang transform was performed on the ice-induced vibration respond signals of structure,the Hilbert energy marginal spectrum of the signals were obtained.Finally,the damage characteristic index was constructed by the variable quantity of Hilbert energy marginal spectrum between healthy structure and damaged structure,the performance of damage identification which was analyzed and compared under different damage locations and damage degrees.The effects of ice velocity and ice thickness on the damage characteristic index and advantages of identify structural damage by ice vibration response were also discussed.Finally,aiming at the risk prediction method of offshore platform ice vibration,A large number of meteorological,hydrological and ice vibration response information were obtained from the field monitoring of the offshore platform in winter,multi-source information fusion is based on the ice vibration hazard classification mechanism,a BP neural network model for ice vibration risk prediction of offshore platform based on genetic algorithm is established,the prediction accuracy of the prediction model is improved obviously.The above research work can provide reference for structural health monitoring of offshore platforms in cold regions.