Study On The Intelligent Damage Detection Of FPSO Soft Yoke Single Point Mooring System

Floating Production Storage and Offloading(FPSO)generally has run on the sea for more than ten years,and permanent single point mooring is the mainly mooring methods in Bohai Sea in China.The single point mooring system on the sea is also subject to the action of FPSO hull in addition to complex environmental loads such as wind and waves,especially at the hinge points.It is difficult for the human to do damage detection by the eye.Structural feature extraction and damage identification based on environmental loads both assume that any structure subjects to a broadband signal.However,the assumptions are difficult to satisfy some structures like marine floating platforms.When non-stationary marine environmental loads act on nonlinear ocean structures,on the other hand,the inherent features of nonlinear ocean structures in different time periods are different,indicating that long-term fixed-point monitoring of the structure is required.According to the features of FPSO soft yoke single point mooring system and the shortcomings in structural monitoring,the multi-coherence function has been applied in this paper to improve the feature extraction method,and the machine learning method has also been used to classify parameters intelligently.Simulations are carried out with the help of software.The raw data of the single point mooring system in site is also used for analysis and verification.The main work is as follows:1.The second-order nonlinear system simulation is carried out by Matlab/Simulink,and the input and output data of nonlinear system with different parameter settings are obtained.The feature extraction is performed by using the coherent function,and the feature classification is performed by SVM(Support Vector Machine)and BP(Back Propagation)classifier.2.The dynamics software ADAMS is used to simulate the dynamics of the soft yoke single point mooring system.The input and output data of this system under different structural parameters are obtained.The feature extraction is performed by using the multi-coherence function.3.Using the multi-coherence function to extract and characterize the input and output of the soft yoke single point mooring system before and after the damage in the Bohai Sea.The FPSO hull motion is used as input,and the response of the mooring leg of soft yoke single point mooring system is used as an output.The features between the input and output of the structure is extracted by the multi-coherence function.The results show that this method can identify the change of the inherent property of the soft yoke single point mooring system,that is,the multicoherence function can be applied for the structural damage detection.The method combined intelligent feature extraction with damage identification based on multi-coherence function and machine learning classification proposed in this paper avoids the assumption of environmental load.The accuracy of the method is verified in the above three nonlinear systems.When the linear system parameters change,that is,when the damages happened,the change can be identified from the change law of the features of the multicoherence function.The accuracy of the structural damage identification results is high,which verifies the effectiveness of the method.It provides a new method for the health monitoring and damage identification of soft yoke single point mooring system,and it provides a theoretical reference for the structural monitoring technology of multi-body structures with input and output such as soft yoke single point mooring system.