Development And Experimental Research Of Intelligent Early Warning System For Vibration Monitoring Of Offshore Platform

The offshore platform is the main facility for the exploration and development of offshore oil and gas resources.The long-term structural health monitoring system arranged on the offshore platform can help managers master the overall structural parameters and damage evolution laws of the platform in real time,which is an effective means to ensure the healthy service of the platform.The purpose of this article is to build an intelligent vibration monitoring and early warning system for the offshore platform based on ambient excitation.The system performs real-time preprocessing,modal parameter identification,and damage warning on monitoring data of offshore platforms,providing reference for future research on structural integrity management of offshore platform.Taking the laboratory offshore platform model as the research object,a data acquisition system is established to collect real-time vibration response data of the offshore platform model under environmental excitation.Preliminary analysis is conducted on the real-time monitoring data.Based on the laboratory monitoring data,signal preprocessing methods are studied.Finally,the processing effect of the signal preprocessing method is verified using indoor offshore platform model data.The results indicate that preprocessing methods such as trend term removal,data smoothing,and variational modal decomposition can effectively remove trend terms,high-frequency noise,and modal peak to peak noise from the original data.To achieve automatic modal parameter identification of offshore platform structure,a modal parameter identification method based on improved empirical wavelet transform is proposed.The sliding data window method is used to segment the real-time data stream.The singular value spectrum of the signal is calculated by singular value decomposition,and the scale space algorithm is introduced to determine the modal peak value and segmentation boundary.Combined with the random decrement technology and Hilbert transform,the automatic identification of modal parameters is realized.Construct simulation signals and ASCE Benchmark model signals,and compare and verify them with the analysis results of empirical wavelet transform and autoregressive empirical wavelet transform.Finally,this method is applied to modal parameter identification of laboratory offshore platform model and on-site offshore platform.The results show that this method can adaptively determine the spectral segmentation boundary,has fast calculation speed,and has high recognition accuracy for the frequency and damping ratio of the structure.Further propose a structural damage warning method for offshore platform based on cointegration and Kalman filtering.Construct a cointegration equation using a sequence of modal parameters under normal structural conditions and calculate cointegration residuals.Determine the warning threshold based on the distribution of cointegration residuals,calculate the cointegration residuals under real-time monitoring conditions,and use Kalman filtering algorithm to perform real-time optimal estimation of the cointegration residuals under monitoring conditions.Combined with the warning threshold,achieve real-time warning.Based on the laboratory offshore platform model,construct data under different working conditions for structural damage warning verification;Based on actual offshore platform data,compare the warning effects before and after Kalman filter optimization.The results show that this method can achieve real-time damage warning,and after optimizing the cointegration residual using Kalman filtering,it can reduce the problem of false alarms in warning.To achieve the engineering application of vibration monitoring and intelligent warning algorithms,the software subsystem was developed based on the Lab VIEW platform.Adopting a "top-down" design approach,the planning of each functional module of the software and the layering of subprograms are achieved,and the functional areas of the software interface are planned.The mixed programming method of MATLAB and Lab VIEW is used to achieve the functions of the main modules such as signal preprocessing,modal parameter identification,intelligent damage warning,etc.Finally,the preparation of other modules is completed.Combining hardware subsystems,build a complete intelligent early warning system for vibration monitoring on offshore platform,and apply it to laboratory offshore platform.The results indicate that the entire monitoring system has strong stability,user-friendly interface,and convenient operation,and can achieve real-time monitoring,automatic modal recognition,and real-time warning of laboratory offshore platform models.It has certain practical engineering application value.