APPLICATION OF THE KERNEL DISCRIMINANT ANALYSIS TO CONDITION IDENTIFICATION FOR SUBMARINE FREE SPANNING PIPELINE
Free spanning pipeline that is suspended between two points on an uneven seafloor, its’ condition variations, such as soil property variation, flow velocity change of currents and waves, axial force alternation and span length variation, directly affect work performance of whole submarine pipeline system. But until now few researches have been focus on condition identification for free span (CIFS). The problem of CIFS has been studied in this paper. The ill-posedness of CIFS is analyzed firstly in detail. Afterwards, this paper presents application of the nonlinear kernel discirminant analysis (KDA) for solving the problem of CIFS. Then the feature parameters are defined in terms of internal constructional characteristics of natural frequencies. Based on the proposed condition vectors consisting of feature parameters, such as natural frequencies, normalized frequencies and frequency change ratios, the KDA can characterize the difference of each condition variation process and implement successfully the objective of classification. Finally, the validity of the KDA approach is evaluated by case study. The analytical results have demonstrated that the proposed approach can identify effectively each condition of free span when condition variation occurs even if under measurement noise. And it is also attractive in the field of engineering application
free span condition identification kernel discriminant analysis (KDA) span length
Jiashun Hu Xin Feng Jing Zhou
State Key Laboratory of Coastal and Offshore Engineering,Dalian University of Technology,Dalian 116023,China
国际会议
第六届国际振动工程会议(The 6th International Conference on Vibration Engineering)(ICVE’ 2008)
大连
英文
2008-06-04（万方平台首次上网日期，不代表论文的发表时间）