考虑环境因素影响的海洋平台结构损伤检测研究

Structural damage detection of offshore platforms considering environmental variations

海洋平台长期服役于恶劣的海洋环境中,结构损伤逐渐累积。海洋环境因素,例如温度、海生物附着、桩基冲刷等,会掩盖真实损伤对结构动力特性的影响,导致基于动力特性的损伤检测方法容易出现损伤误判。将协整分析技术应用于海洋平台结构的损伤判定中,通过对结构频率序列进行协整分析,消除环境因素干扰,提取能够表达结构真实动力特性的协整残差,而后依据X-bar控制图进行损伤判定。以最为普遍的多变性环境因素温度为例,同时考虑空气、海水和泥土温度变化的影响,探究协整方法对消除这些变化的有效性,并与常用的主成分分析(PCA)方法进行对比。研究结果表明,两种方法均能有效地消除温度变化的影响,实现海洋平台结构损伤的准确判定。小噪声水平下,协整分析的稳定性更好,随着噪声水平的增加,主成分分析具有更高的噪声鲁棒性。此外,协整分析在基准数据较少的情况下具有更低的误判率。

Offshore platforms continually serve in atrocious ocean environment,imposing the occurrence and growth of structural damage.Environmental variations,such as temperature,marine growth,foundation scour,and etc.,tend to mask the changes of structural dynamic features caused by damage.To this end,the vibration-based methods cannot be used to identify the existence of the damage.In this paper,a cointegration analysis-based method was proposed for damage detection of offshore platforms and the cointegration residuals of natural frequencies extracted by the cointegration analysis were used to remove the environmental variations.Further,the X-bar chart was used to determine the occurrence of damage.The influence of temperature,including air temperature,water temperature and seabed mud temperature was simultaneously considered because temperature is the most common condition that changes the dynamic features of structures.The performance of cointegration method was explored and compared with the principle component analysis,which has been constantly used for damage detection under environmental variations.Results show that both methods can successfully remove the temperature influence and accurately identify the damage.The cointegration performs more robust with a low noise level while the principle analysis behaves well with a high noise level.Besides,cointegration has a lower misjudgment rate in the case of small baseline sample.