摘要
针对某水下结构的抗水下爆炸试验数据,从爆炸振动特征和结构自身特性这两个方面出发,基于小波包技术选择加速度峰值、主振频带能量和自振频带能量作为特征量。由于这些特征量之间的强相关性,利用正交旋转法的因子分析提取出代表结构毁伤大部分信息的本征特征量。然后分别利用原始特征量数据和本征特征量数据进行结构毁伤判别,两组数据的判别结果与实验结果基本一致,证明了本征特征量提取的合理性。最后给出了结构毁伤距离曲线。
Abstract ：Based on the experiment data of a underwater structure subjected to underwater explosion and wavelet packer analysis theory, Nine characteristic values were taken into account from the aspects of explo- sion vibration and structure dynamic characteristics, such as the peak acceleration, the energy on dominant frequency bands and natural frequency bands of interior device. Because of the strong correlation between these characteristic values, the intrinsic characteristic values affecting the most target damage information were extracted by using the orthogonal rotating of factor analysis. Results show that the prediction results are basically consistent with the experimental results by distingishing target damage and prove that the choice of intrinsic characteristic values is reasonable. Finally, the structure＇s damage distance curve is obtained.
出处
《船舶力学》
EI
CSCD
北大核心
2014年第1期90-97,共8页
Journal of Ship Mechanics
基金
国家自然科学基金(50775218)
海军工程大学博士生创新基金(HGBSJJ2011009)
关键词
水下爆炸
毁伤评估
小波包
因子分析
费歇尔判别法
underwater explosion
damage assessment
wavelet packet
factor analysis
Fisher discriminant
