基于EEMD的立管涡激振动响应最优降噪光滑模型参数识别研究

A study on parameter identification of optimal noise reduction smooth model for vortex-induced vibration response of riser based on EEMD

深水海洋立管因受到海洋环境的影响产生涡激振动现象,长期的涡激振动易导致立管产生疲劳破坏。为了获得真实的涡激振动特性,进行海洋立管涡激振动试验并对测试信号进行总体平均经验模态(EEMD)分解。基于分解得到的本征模态函数建立滤波算法,考虑曲线光滑度和重构信号与实测信号相似度,确立最优降噪光滑模型区间,依据存在度确定有效本征模态函数以及最优降噪光滑模型,优化目标函数。从相关度、频谱分析、能量谱分析3个角度出发进行合成信号仿真识别,并验证了算法的准确性。结果表明:构建的最优降噪光滑模型得到的有效本征模态函数,与原始信号频率成分相关系数最高,达到0.96以上,且频率对应相等;其能量占原始信号总能量的97.98%。涡激振动试验结果表明,建立的最优降噪光滑模型确定的有效本征模态函数包含顺流向、横流向振动信息,且顺流向主频是横流向频率的2倍。

In order to obtain the real characteristics of vortex-induced vibration,the overall ensemble empirical mode decomposition(EEMD)of the signals was carried out.A filtering algorithm was established using the decomposed intrinsic mode function.Considering the smoothness of the curve and the similarity between the reconstructed signal and the measured signal,the optimal noise reduction smooth model interval was established.The effective intrinsic mode function was determined according to the degree of existence.The synthetic signal was simulated and identified from the perspectives of correlation degree,spectrum analysis and energy spectrum analysis.The accuracy of the algorithm is verified from correlation degree,spectrum analysis and energy spectrum analysis.Results show that the optimized IMFs contain the information of vortex-induced vibration,and the main frequency in the downstream direction is twice as much as that in the transverse direction.