Research On Health Status Analysis Of Offshore Platforms Based On MTS

As an important offshore oil development, the offshore platform equipment has dual characteristics of a large weight and complex structure. The health status of this kind of large project structure is the key to the development of offshore oil fields. In recent years, the occurrence of many offshore platforms due to fatigue damage has been seen caused by the collapse. For the offshore platform, the damage during service period in offshore platform structure is inevitable. Health monitoring and damage diagnosis of the offshore platform has become an important issue recently. Using vibration measurement and analysis technology to identify the platform damage, it is not only helpful to make a general evaluation for the whole structure, but also offer necessary data for the system identification.The existing analysis methods of offshore platform status is mainly according to the offshore platform structural or mechanical angle, and these kind of methods for offshore platform which needs large mechanical operation of long-term accumulation of work states analysis is very effective, but the health monitoring of offshore platform for real-time analysis including establishing an evaluation system to evaluate the operation environment has not been solved in fact.This paper introduces a new method of pattern recognition of multiple quantitative strategy into analyzing the data of offshore platform by using the method of MTS (Mahalanobis-Taguchi System). Setting the "No.102" offshore platform data as an example, this dissertation sets up an analysis system of platform health status. By using the MATLAB, SPSS, MINITAB analysis software, normal and abnormal state data will be extracted to samples classify. The orthogonal table is built to screen variables in this system. And the analysis of the characteristics of the signal-to-noise ratio results provides the optimization of the data set. Finally, compared the two threshold methods (ROC curve analysis and Fisher discrimination analysis to determine the critical value of the mahalanobis distance calculation), the research obtain the better way to make result more accurate. As a new method for the offshore platform health analysis, this paper provides a new method for pattern recognition and prediction for unknown samples in the future.